Paper/Power electronics

[논문리뷰] Optimal Brake Allocation in Electric Vehicles for Maximizing Energy Harvesting During Braking <3>

얼죽아여뜨샤 2024. 4. 5. 13:49

0. 원문

Optimal_Brake_Allocation_in_Electric_Vehicles_for_Maximizing_Energy_Harvesting_During_Braking.pdf
2.84MB

 

1. 내용

(0) Abstract

This article proposes a novel approach to efficiently distribute braking force of an electric vehicle (EV) between friction and regenerative braking with an ultimate goal of maximizing harvested energy during braking.

The regenerative braking performance of an EV depends on various factors influenced by the driver behavior and driving conditions, which are challenging to measure or predict in real-time.

이 논문은 전기 자동차(EV)의 제동 힘을 마찰 및 재생 제동 사이에 효율적으로 분배하기 위한 새로운 방법을 제안합니다. 이 방법의 궁극적인 목표는 제동 중 수확된 에너지를 극대화하는 것입니다. 

EV의 재생 제동 성능은 운전자 행동 및 주행 조건에 영향을 받는 다양한 요인에 따라 결정되며, 실시간으로 측정하거나 예측하기가 어려운 도전적인 문제입니다. 

In the proposed method, the performance map of the traction motor (TM) and its controller is used to define a boundary in which blending of regenerative and friction braking is performed with the goal of maximizing recaptured energy through the regenerative braking process.

제안된 방법에서는 구동 모터(TM) 및 해당 컨트롤러의 성능 지도를 사용하여 재생 및 마찰 제동의 혼합이 수행되는 경계를 정의하고, 재생 제동 프로세스를 통해 회수된 에너지를 극대화하는 것을 목표로 합니다. 

Traction motor(TM)
: 전기의 힘으로 회전하여 바퀴를 구동하는 모터 장치이다. 전기자동차 구동모터는 전기를 이용하여 구동력을 발생하는 전장품으로, 모터 축에 감속기를 연결하여 적절한 토크를 바퀴에 전달하여 차량을 구동하는 부품이다. 
구동모터는 전기에너지로부터 전자계를 매개체로 하여 기계에너지로 변환을 통해 구동력을 발생시키기 장치는 회전기기 또는 선형기기이다. 

 

The performance, effectiveness, and robustness of the proposed strategy are validated through a hardware-in-the-loop (HIL) experimental testbed for a predetermined drive cycle of Urban Dynamometer Driving Schedule (UDDS).

제안된 전략의 성능, 효과 및 견고성은 Urban Dynamometer Driving Schedule (UDDS)의 사전 결정된 운전 주기를 위한 하드웨어 인 더 루프(HIL) 실험 테스트베드를 통해 유효성이 검증됩니다. 

Hardware-in-the-loop(HIL)
: HIL(Hardware-in-the-Loop) 시뮬레이션은 제어할 물리적 시스템을 나타내는 가상 실시간 환경을 생성하여 의도한 대상 컨트롤러에서 실행되는 제어 알고리즘을 검증하는 기술입니다. HIL은 실제 프로토타입 없이 제어 알고리즘의 동작을 테스트하는 데 도움이 됩니다.


Urban Dynamometer Driving Schedule(UDDS)
: UDDS 사이클은 테스트 차량의 속도를 높였다가 다시 0으로 낮추는 반복 주행하며, 이는 도시 연비를 측정하는데 사용된다. HWFET 사이클은 차량의 속도를 높인 다음 30-60mph 범위의 다양한 속도 사이로 주행하며, 이는 고속도로 연비를 측정하는데 사용된다.
Urban Dynamometer Driving Schedule (UDDS)는 자동차 및 차량의 배기가스 및 연료 소비를 평가하기 위해 사용되는 표준 시험 절차 중 하나입니다. UDDS는 실제 도시 운전 패턴을 시뮬레이션하여 자동차의 연비와 배기가스 배출량을 측정합니다. 이 시험은 EPA (미국 환경보호국)의 도시 운전 사이클 중 하나로 사용되며, 자동차 및 차량의 연비 등급을 결정하는 데 중요한 역할을 합니다. UDDS는 도시에서의 주행을 모방하여 주로 저속 및 중간 속도에서의 운전 패턴을 포함하고 있습니다.

 

It is shown that using the proposed method, the amount of recaptured energy through the regenerative braking process can significantly increase compared to constant or variable boundary methods using a weight factor for brake distribution.

제안된 방법을 사용하면, 재생 제동 프로세스를 통해 회수된 에너지량이 상수 또는 가변 경계 방법을 사용하는 것과 비교하여 상당히 증가함을 보여줍니다.

 

#Keyword : Brake allocation, electric vehicle (EV), hardwarein- the-loop (HIL), low-speed boundary, performance map,
regenerative braking, test bench, traction motor (TM).

 

(1). Introduction

WITH the evolution of new technologies such as artificial intelligence and growing interest in environmental conservation and sustainability, autonomous and electrified vehicles have become the foremost focus of automotive industries and research institutions [1].

Electric vehicle (EV), hybrid EV (HEV), and plug-in HEV (PHEV) configurations have been the prime focus of this evolution due to their immense impact on reducing fossil-fuel based energy consumption and greenhouse emissions [2]–[4].

Although HEVs and PHEVs have the advantage of a longer driving range compared to EVs due to their secondary fossil-fuel based energy source [5], EVs are considered the cleanest alternatives to conventional vehicles [6] and thus have the highest impact on transitioning to a sustainable transportation future.

However, certain characteristics of EVs compared to those of conventional vehicles and HEVs/PHEVs still need to be improved in order for them to be considered as market leaders of the transportation sector [6].

Even though certain challenges of EVs related to traction motor (TM) and inverter efficiency, energy management strategies, and the associated control systems have been significantly alleviated, EVs have yet to gain wide acceptance due to their limited driving range [7], [8].

As a result, large-scale commercialization of EVs is hindered [7].

신기술인 인공 지능과 환경 보존 및 지속 가능성에 대한 관심이 증가함에 따라 자율주행 및 전기화된 차량이 자동차 산업 및 연구 기관의 주요 관심사로 부상했습니다. 

전기 자동차(EV), 하이브리드 EV(HEV) 및 플러그인 HEV(PHEV) 구성이 환경 보존 및 온실 가스 배출량 감소에 미치는 엄청난 영향 때문에 이러한 진화의 주요 관심사가 되었습니다. 

비록 HEV 및 PHEV는 이차 화석 연료 기반 에너지원으로 인해 EV에 비해 긴 주행 거리의 장점을 가지지만, EV는 기존 차량에 대한 가장 깨끗한 대안으로 간주되어 지속 가능한 교통의 미래로의 전환에 가장 큰 영향을 미칩니다. 

그러나 EV의 특성 중 일부는 여전히 기존 차량 및 HEV/PHEV와 비교하여 개선되어야 합니다.

전기 자동차의 추진 모터(TM) 및 인버터 효율, 에너지 관리 전략 및 관련된 제어 시스템과 관련된 특정 과제가 상당히 완화되었지만, 제한된 주행 거리로 인해 EV는 아직까지 널리 받아들여지지 않고 있습니다.

결과적으로, 대규모로 전기 자동차를 상용화하는 것이 제약됩니다.

 

Numerous research studies in both industry and academia are being conducted to improve the driving range and overall efficiency of EVs [9]–[11].

Detailed investigation has been made to analyze the energy consumption distribution in different driving cycles [7], [12].

Based on the results, in general, about 42% of the energy used in a typical vehicle is consumed for propelling the vehicle, 25% is wasted in the form of heat during vehicle deceleration and braking, 23% is wasted by air drag, and 10% is used as other forms [7].

Since the required energy to propel the vehicle cannot be significantly reduced [13], and the energy wasted by air drag depends on the shape of the vehicle, extracting the mechanical energy of the vehicle during braking instances has become the focus of researchers in order to make future EVs more efficient.

If the mechanical energy of an EV is efficiently recaptured, transformed into electrical energy, and stored in the battery during braking instances, the driving range per battery charge and the overall efficiency will significantly improve [7], [14].

EVs have the advantage of regenerative braking in terms of energy savings, especially in metropolitan areas with frequent stops [15], [16].

The action of regenerative braking mainly employs the back electromotive force (EMF) of the TM during braking such that the back EMF is regarded as a voltage source to recharge the battery [10].

In this process, controlled switching of the inverter is required to efficiently recapture the mechanical energy of the vehicle [10].

산업 및 학계에서는 EV의 주행 거리 및 전반적인 효율성을 향상시키기 위한 다양한 연구가 진행되고 있습니다. 

다양한 주행 주기에서의 에너지 소비 분포를 분석하기 위해 상세한 조사가 수행되었습니다. 

이러한 결과를 기반으로 일반적으로 전형적인 차량에서 사용되는 에너지 중 약 42%는 차량을 추진하는 데 사용되고, 25%는 차량 감속 및 제동 중에 열 형태로 낭비되며, 23%는 공기 저항에 의해 낭비되고 있으며, 10%는 기타 형태로 사용됩니다. 

차량을 추진하는 데 필요한 에너지를 크게 줄일 수 없으므로, 공기 저항에 의한 에너지 낭비는 차량의 모양에 따라 달라집니다. 따라서 미래의 EV가 더 효율적이 되기 위해 차량의 기계 에너지를 제동 시간에 효율적으로 회수하는 것이 연구자들의 주요 관심사가 되었습니다. 

EV의 기계 에너지가 제동 시간에 효율적으로 회수되고 전기 에너지로 변환되며 배터리에 저장된다면, 배터리 충전 당 주행 거리와 전반적인 효율이 크게 향상됩니다. 

특히 자주 멈추는 대도시 지역에서는 에너지 절약 측면에서 회생 제동의 이점을 가지고 있습니다. 

회생 제동의 작동은 주로 제동 중에 TM의 역 전기력(EMF)을 활용하므로 역 EMF가 배터리를 재충전할 전압원으로 간주됩니다. 

이 과정에서 차량의 기계 에너지를 효율적으로 회수하기 위해 인버터의 제어된 스위칭이 필요합니다.

 

Theoretically, a fully electrified braking system for EVs, in which all the braking is realized by regenerative braking, is feasible [17].

However, due to limitations on the produced resistive torque by the TM under heavy braking circumstances [18], and the possibility of failure in the electrical system that may lead to failure of the regenerative braking system [19], coexistence of conventional friction-based brakes is inevitable [20], [21].

Therefore, an extremely important consideration regarding EV brake system design is to effectively allocate braking energy between friction and regenerative braking not only to maintain the vehicle stability/controllability, but also to maximize the amount of absorbed energy [22].

Ideally, regenerative braking should be utilized to recapture as much energy as possible without compromising the vehicle stability and controllability [22].

이론적으로, 모든 제동이 회생 제동으로 실현되는 전적으로 전기화된 EV용 제동 시스템은 가능합니다.
그러나, TM이 강한 제동 상황에서 생산하는 저항 토크에 제한이 있고 [18], 전기 시스템의 고장 가능성이 회생 제동 시스템의 고장으로 이어질 수 있으므로 [19], 전통적인 마찰 제동 장치의 공존은 불가피합니다 [20], [21].
따라서 EV 제동 시스템 설계에 대한 매우 중요한 고려 사항은 차량의 안정성/조종성을 유지하는 데 그치지 않고 획득된 에너지량을 최대화하기 위해 마찰과 회생 제동 간의 제동 에너지를 효과적으로 할당하는 것입니다 [22].
이상적으로는, 회생 제동은 차량의 안정성과 조종성을 저해하지 않으면서 가능한 한 많은 에너지를 회수하기 위해 활용되어야 합니다 [22].

 

Taking into account the TM constraints, battery capacity limitation, and the fact that in an EV only the energy extracted through the drive axle is effective in the regenerative braking process, the brake controller is responsible for accurate blending of regenerative and friction-based brakes [22].

Regarding TM constraints, two main limitations should be considered while designing the brake controller [22].

The first limitation is the maximum regenerative braking capability, which stems from the braking torque capability of the TM while operating in regenerative braking mode [23].

The other limitation, which is the main scope of this paper, is the inability of the TM and the controller to harvest the mechanical energy of the vehicle at very low speeds even though the TM is operated as a generator [24].

At low speeds, the induced back EMF, which is proportional to the motor rotational speed, is not sufficient to overcome losses related to regenerative braking, resulting in energy extraction from the battery instead of recharging it [10], [25].

TM의 제약, 배터리 용량 제한, 그리고 EV에서 회생 제동 프로세스에서 구동 축을 통해 추출된 에너지만 효과적이라는 사실을 고려할 때, 제동 컨트롤러는 회생 제동과 마찰 제동을 정확하게 혼합하는 것을 담당합니다 [22]. 

TM 제약 사항에 대해 제동 컨트롤러를 설계할 때 고려해야 할 두 가지 주요 제한 사항이 있습니다 [22]. 

첫 번째 제한 사항은 TM이 회생 제동 모드에서 작동하는 동안의 제동 토크 능력에서 비롯된 최대 회생 제동 능력입니다 [23]. 

다른 제한 사항은 본 논문의 주요 범위인데, 이는 TM과 컨트롤러가 저속에서 차량의 기계 에너지를 회수하지 못하는 능력입니다. 

비록 TM이 발전기로 작동하더라도 저속에서는 모터 회전 속도와 비례하는 유도된 역 전기력이 회생 제동과 관련된 손실을 극복하기에 충분하지 않기 때문에, 에너지를 배터리에서 추출하여 다시 충전하지 못합니다 [10], [25].

 

One approach to improve energy extraction during regenerative braking process is to consider a low-speed boundary, under which regenerative braking is no longer efficient [15].

The effect of low-speed boundary on optimum allocation of regenerative and friction-based brakes has been comprehensively studied in the literature [26]–[33].

In [26]–[28], a fixed low-speed boundary is considered in the brake controller during the blending of regenerative and friction-based brakes.

However, as shown in [29], this low-speed boundary varies under different vehicle/road conditions.

Thus, it should be considered variable as opposed to a fixed boundary and should be determined based on the vehicle operating points and road conditions.

회생 제동 과정에서 에너지 추출을 개선하는 한 가지 방법은 회생 제동이 더 이상 효율적이지 않은 저속 경계를 고려하는 것입니다 [15]. 

저속 경계가 회생 및 마찰 제동의 최적 할당에 미치는 영향은 문헌에서 철저하게 연구되었습니다 [26]–[33]. 

[26]–[28]에서는 회생 및 마찰 제동의 혼합 중 제동 컨트롤러에서 고정된 저속 경계가 고려됩니다. 

그러나 [29]에서 보듯이, 이 저속 경계는 차량/도로 조건에 따라 다르게 변합니다. 

따라서 이는 고정된 경계가 아닌 가변적으로 고려되어야 하며, 차량 운전 점과 도로 조건에 기반하여 결정되어야 합니다. 

In [30], a fuzzy control theory is applied to a brake force distribution of a front-wheel drive EV. In the presented fuzzy control strategy, due to low efficiency of regenerative braking process at lowspeeds, regenerative braking is given a lower proportion of the total braking force when the speed is low.

[30]에서는 전륜 구동 EV의 제동 힘 분배에 퍼지 제어 이론이 적용됩니다. 

제시된 퍼지 제어 전략에서 저속에서 회생 제동의 효율이 낮기 때문에 속도가 낮을 때 전체 제동 힘 중 회생 제동에 낮은 비율이 할당됩니다. 

Furthermore, in [31], a weight factor control strategy is applied to the brake force distribution of a front-wheel drive EV.

In this strategy, braking force is distributed between friction and regenerative braking such that regenerative braking share is increased with an increase in the speed.

또한, [31]에서는 전륜 구동 EV의 제동 힘 분배에 가중치 제어 전략이 적용됩니다. 

이 전략에서는 회생 제동과 마찰 제동 사이의 제동력이 속도가 증가함에 따라 증가하도록 조정됩니다. 

In both [30] and [31], the calculated low-speed regenerative braking boundary is dependent on a limited number of influencing factors associated with vehicle and road operating conditions.

However, many other factors including but not limited to weather conditions, driver behavior, and road slope angle can have an influence on the displacement of low-speed boundary.

[30]과 [31] 모두 계산된 저속 회생 제동 경계는 차량 및 도로 운영 조건과 관련된 영향 요소의 제한된 수에 의존합니다. 

그러나 기상 조건, 운전자 행동 및 도로 경사 각도와 같은 다른 많은 요소들도 저속 경계의 변위에 영향을 줄 수 있습니다. 

In [32] and [33], the brake force distribution is implemented in the brake controller using different artificial neural network control strategies.

In these methods, the training dataset should contain sufficient information about regenerative braking and friction braking forces for different operating conditions.

In other words, the selection of the drive cycle, which is used in training procedure to obtain the training dataset, is very important for simulating sufficient braking scenarios and the performance of the brake controller is highly influenced by the drive cycles.

[32]와 [33]에서는 인공 신경망 제어 전략을 사용하여 제동 컨트롤러에서 제동 힘 분배가 구현됩니다. 

이러한 방법에서는 훈련 데이터 집합에 다양한 운영 조건에 대한 회생 제동 및 마찰 제동 힘에 대한 충분한 정보가 포함되어야 합니다. 

다시 말해, 훈련 절차에서 사용되는 주행 주기 선택은 충분한 제동 시나리오를 시뮬레이션하고 제동 컨트롤러의 성능이 주행 주기에 크게 영향을 받습니다.

 

In [15], a dynamically changing low-speed boundary is considered in the brake controller by monitoring TM controller’s dc link current.

However, in this approach any inaccuracy in dc link current measurement at low currents will directly influence the outcome and impair the accuracy of detecting zero current crossing points.

Considering that the existing current sensors in EVs do not provide this level of accuracy for low currents, the transition between regenerative and friction braking may not occur at its optimum point.

In addition, implementing the blending between regenerative and friction braking based on the method used in [15] is associated with delays in measuring real-time current and feeding it back to the brake controller for analysis.

[15]에서는 TM 컨트롤러의 직류 링크 전류를 모니터링하여 제동 컨트롤러에서 동적으로 변하는 저속 경계가 고려됩니다. 

그러나 이 방법에서 낮은 전류에서의 직류 링크 전류 측정의 정확도에 대한 어떤 부정확성도 결과에 직접적으로 영향을 미치고 영점 통과점을 감지하는 정확도를 저해할 수 있습니다. 

EV의 기존 전류 센서가 낮은 전류에 대해 이 수준의 정확도를 제공하지 않는다고 고려할 때, 회생 및 마찰 제동 간의 전환은 최적의 지점에서 발생하지 않을 수 있습니다. 

또한, [15]에서 사용된 방법을 기반으로 회생 및 마찰 제동 간의 혼합을 구현하는 것은 실시간 전류를 측정하고 제동 컨트롤러에 실시간으로 피드백하여 분석하는 데 시간이 걸리는 지연과 관련이 있습니다. 

In this paper, a novel approach is proposed that relies on the TM regenerative braking performance map to identify the low-speed boundary and achieve optimum brake control in EVs.

The main technical contributions of this paper are: 1) to outline the operating principle of regenerative braking and mathematically express the limitations of regenerative braking at low speed, and 2) to propose an effective approach based on the TM regenerative braking performance map that can achieve optimum blending of regenerative and friction-based brakes and improve energy harvesting during braking.

본 논문에서는 TM 회생 제동 성능 지도를 기반으로 낮은 속도 경계를 식별하고 EV에서 최적의 제동 제어를 달성하는 신규 접근 방법을 제안합니다. 

이 논문의 주요 기술 기여는 다음과 같습니다: 1) 회생 제동의 작동 원리를 개요하고 낮은 속도에서의 회생 제동의 제한을 수학적으로 표현하는 것, 2) TM 회생 제동 성능 지도를 기반으로 회생 및 마찰 제동의 최적 혼합을 달성하고 제동 중 에너지 수확을 개선할 수 있는 효과적인 접근 방법을 제안하는 것입니다.

 

The remainder of this paper is organized as follows.

In Section II, operating principle of regenerative braking and its low-speed limitation are presented.

In Section III, theTMregenerative braking performance map is analyzed and experimentally obtained for further use in the brake controller, and in Section IV, the proposedEVbrake controller design is explained in detail.

In Section V, case studies are carried out on a hardware-in-the-loop (HIL) test bench, and experimental results are presented to validate the effectiveness of the proposed approach. Finally, conclusions are drawn in Section VI.

이 논문의 나머지는 다음과 같이 구성되어 있습니다. 

II장에서 회생 제동의 작동 원리와 저속 제한에 대해 소개합니다. 

III장에서는 TM 회생 제동 성능 지도를 분석하고 제동 컨트롤러에서 사용하기 위해 실험적으로 얻은 내용을 설명합니다. 

IV장에서는 제안된 EV 제동 컨트롤러 설계를 자세히 설명합니다. 

V장에서는 하드웨어 인 더 루프(HIL) 테스트 벤치에서 사례 연구를 수행하고 제안된 접근 방법의 효과를 검증하는 실험 결과를 제시합니다. 

마지막으로, 결론을 VI장에서 도출합니다.

 

(2) Regenerative Breaking and Low-Speed Limitation

This work focuses on a three-phase motor with sinusoidal back EMF as theTMof an EV.

Fig. 1 shows an equivalent circuit of a typical three-phase motor and its inverter that is connected to a dc voltage source representing the EV battery.

이 연구는 EV의 TM으로 Sinusoidal Back EMF를 갖는 3상 모터에 중점을 두고 있습니다. 

그림 1은 전형적인 3상 모터의 등가 회로와 EV 배터리를 나타내는 DC 전압원에 연결된 인버터를 보여줍니다.

 

In Fig. 1, R and L represent the armature resistance and inductance of the motor, respectively, while ea, eb, and ec are the armature back EMFs and ia, ib, and ic are the armature currents of phase a, b, and c, respectively. 

Moreover, in this model,RLoss represents the cable losses as well as the losses due to internal resistance of the battery. 

For simplicity, inverter losses are not considered in this study.

그림 1에서 R 및 L은 모터의 armature 저항과 인덕턴스를 각각 나타내며, ea, eb 및 ec는 각각 상 a, b 및 c의 armature 역 기전력 기준이고, ia, ib 및 ic는 각각 상 a, b 및 c의 armature 전류를 나타냅니다. 

또한, 이 모델에서 RLoss는 케이블 손실과 배터리의 내부 저항에 따른 손실을 나타냅니다. 

간단히 하기 위해, 이 연구에서는 인버터 손실을 고려하지 않습니다.

 

Generally, a special switching strategy is required for the inverter to recapture the mechanical energy of the vehicle while decelerating.

A typical switching strategy discussed in [10] is utilized in this section to study regenerative braking limitations at low speed. 

[10] : A Cost-Effective Method of Electric Brake With Energy Regeneration for Electric Vehicles
위 논문에 이 논문에서 사용된 식들의 원본이 있음, 꼭 참고!

 

Based on [10], during normal/regenerative braking modes of operation, a complete commutation sequence consists of six interval states.

Since the method to analyze switching patterns of states II-VI is similar to that of state I, the analysis is simplified by only considering state I.

In this regard, the equivalent circuits of state I for both modes of normal and regenerative braking are shown in Figs. 2 and 3, respectively [10].

일반적으로, 차량의 감속 중에 기계 에너지를 회수하기 위해 인버터에 특별한 스위칭 전략이 필요합니다. 

이 섹션에서는 저속에서의 회생 제동 한계를 연구하기 위해 [10]에서 논의된 전형적인 스위칭 전략이 사용됩니다. 

[10]에 따르면, 정상/회생 제동 운전 모드 중에 완전한 전환 순서는 여섯 개의 간격 상태로 구성됩니다. 

상태 II-VI의 전환 패턴을 분석하는 방법이 상태 I와 유사하므로, 분석은 상태 I만을 고려하여 단순화됩니다. 

이와 관련하여, 정상 및 회생 제동 모드의 상태 I의 등가 회로는 각각 그림 2와 그림 3에 나타나 있습니다.

 

As depicted in Fig. 2, the normal mode of operation consists of two modes of conduction and freewheeling. 

In this mode of operation, switch S1 is operated with pulse width modulation (PWM) signals while switch S4 is always ON. 

The conduction mode occurs when both switches are ON at the same time resulting in a current path shown by a solid line (iON) in Fig. 2. 

The freewheeling mode occurs when switch S1 is in OFF state and switch S4 is ON, resulting in a current path shown by a dashed line (iOFF ) in Fig. 2.

그림 2에 표시된 것처럼, 정상 운전 모드는 동작 및 자유회전 두 모드로 구성됩니다. 

이 모드에서 스위치 S1은 펄스 폭 변조(PWM) 신호와 함께 작동되며 스위치 S4는 항상 켜진 상태입니다. 

동작 모드는 두 스위치가 동시에 켜진 상태로 발생하여 그림 2의 실선으로 표시된 전류 경로(iON)를 나타냅니다. 

자유회전 모드는 스위치 S1이 꺼진 상태이고 스위치 S4가 켜진 상태일 때 발생하여 그림 2의 대시선으로 표시된 전류 경로(iOFF)를 나타냅니다.

 

As displayed in Fig. 3, the regenerative mode of operation, in which switches S2 and S3 are controlled with a PWM signal, is also made up of two modes of conduction (both switches are ON with the current path iON) and freewheeling (both switches are OFF with the current path iOFF ).

그림 3에 나타난 바와 같이, 스위치 S2 및 S3이 PWM 신호로 제어되는 회생 모드 운전은 동작(두 스위치가 켜져 있는 상태로 전류 경로 iON) 및 자유회전(두 스위치가 꺼진 상태로 전류 경로 iOFF) 두 가지 모드로 구성됩니다.

 

Regarding the regenerative braking mode, during PWM ON (both switches are ON), the line-to-line inductor voltage (V2L(ON)) is calculated as

회생 제동 모드에 관해서는, PWM ON 상태에서(두 스위치가 모두 켜진 상태) 선간 인덕터 전압 (V$2L(ON)$)은 다음과 같이 계산됩니다.

where Vbatt is the dc voltage of the battery, eab is the line-to-line back EMF generated by the motor, and Vdrop is the voltage drop due to line-to-line equivalent resistance of the motor (2R) and RLoss.

On the other hand, during PWMOFF (both switches are OFF), V2L(OFF) is calculated as

여기서 V$batt$는 배터리의 직류 전압, eab는 모터에 의해 생성된 선간 역전압, V$drop$은 모터의 선간 등가 저항 (2R)과 RLoss로 인한 전압 강하입니다.

반면, PWM OFF (두 스위치가 모두 꺼진 상태) 동안 V$2L(OFF)$은 다음과 같이 계산됩니다.

 

Considering that in steady state operation, the average voltage across the inductor in one cycle is zero, one can obtain

정상 상태에서 한 주기 동안 인덕터에 걸리는 평균 전압이 0이라고 가정하면, 다음이 얻어집니다.

where ΔtON is the duration in which both S2 and S3 switches are ON and ΔtOFF is the duration in which both switches are OFF during one PWM cycle.

Δt$ON$은 S2와 S3 스위치가 모두 켜져 있는 기간이고, Δt$OFF$는 한 PWM 주기 동안 두 스위치가 모두 꺼져 있는 기간입니다.

 

Regarding the regenerative braking mode, the battery energy is delivered to the motor during PWM ON; on the contrary, during PWM OFF, the extracted braking energy is pushed back into the battery. 

This energy can be calculated for each mode as

재생 제동 모드에서, 배터리 에너지는 PWM ON 동안 모터로 공급되고, 반면에 PWM OFF 동안 추출된 제동 에너지는 다시 배터리로 밀어 넣습니다. 

각 모드에 대한 이 에너지는 다음과 같이 계산될 수 있습니다.

 

Based on (3)-(5), the proportion of energy pushed back into the battery (ΔE$OFF$) to the energy extracted from the battery (ΔE$ON$) during regenerative braking can be expressed as

식 (3)~(5)에 따르면, 재생 제동 중 배터리에서 추출된 에너지(ΔE$ON$)에 비해 배터리로 밀어 넣는 에너지(ΔE$OFF$)의 비율은 다음과 같이 표현될 수 있습니다.

 

Based on (6), α is always negative since Δt$OFF$ and Δt$ON$ are always positive.

Also, since eab is always less than Vbatt by design [10], one can conclude that the term (eab − Vbatt − Vdrop) will always have a negative value. 

As long as α < −1, the proportion of energy pushed back into the battery in one cycle will be greater than the extracted energy, which indicates efficient operation of regenerative braking. 

In contrast, α > −1 indicates that the energy pushed back into the battery in one cycle is less than the energy extracted from it and therefore regenerative braking is inefficient. 

Considering that (eab − Vbatt − Vdrop) is always negative, this will only occur in situations where eab < Vdrop. 

Hence, energy extraction at low speeds is dependent on both eab and Vdrop values and regenerative braking process is only effective in charging the battery if the motor line-to-line back EMF is greater than the overall voltage drop.

식 (6)에 따르면 ΔtOFF와 ΔtON이 항상 양수이므로 α는 항상 음수입니다. 

또한, eab가 설계상 Vbatt보다 항상 작기 때문에(eab < Vbatt), (eab - Vbatt - Vdrop) 항목은 항상 음수 값을 가질 것입니다. 

α가 -1보다 작은 동안은 한 주기 동안 배터리로 밀어 넣는 에너지의 비율이 추출된 에너지보다 크기 때문에 재생 제동이 효율적으로 작동함을 나타냅니다. 

반면에 α가 -1보다 크면 한 주기 동안 배터리로 밀어 넣는 에너지가 배터리에서 추출된 에너지보다 작으므로 재생 제동이 비효율적입니다. 

(eab - Vbatt - Vdrop)이 항상 음수임을 고려하면, 이는 eab < Vdrop인 상황에서만 발생합니다.

따라서 저속에서의 에너지 추출은 eab 및 Vdrop 값 모두에 의존하며, 모터 라인 간 back EMF가 전체 전압 강하보다 클 때에만 재생 제동 과정이 배터리를 충전하는 데 효과적임을 의미합니다.

 

Results from this analysis clearly show that the process of regenerative braking leads to energy loss during instances when the voltage drop exceeds the back EMF voltage. 

This phenomenon occurs at low speeds indicating that a low speed boundary should be considered in the brake controller to disable regenerative braking.

Moreover, since the overall voltage drop is associated with current and during regenerative braking this current can fluctuate depending on the required regenerative torque, this low-speed boundary varies dynamically based on different factors as shown in [15].

이 분석 결과에서는 재생 제동 과정이 전압 강하가 백 EMF 전압을 초과하는 경우 에너지 손실을 초래한다는 것을 명백히 보여줍니다.  => eab < Vdrop 인 경우를 뜻하는 듯함

이 현상은 저속에서 발생하며, 이러한 경우에는 저속 경계가 브레이크 컨트롤러에서 재생 제동을 비활성화해야 함을 나타냅니다. 

게다가, 전체 전압 강하는 전류와 관련이 있으며, 재생 제동 중 이 전류는 필요한 재생 토크에 따라 변동할 수 있으므로 이 저속 경계는 다양한 요소에 따라 동적으로 변동합니다([15] 참조).

 

(3) TM Regenerative Braking Performance Map

To experimentally show the regenerative braking limitations of a TMduring low-speed operation, a high-performance permanent magnet synchronous motor (PMSM) connected to a 400-V battery unit is tested while operating as a generator.

The PMSM is operated as a generator for different low speeds and resistive torques such that the resistive torque, shaft speed, and battery current are recorded at every step.

The PMSM is capable of operating in the speed range of 0–7700 rpm, has a maximum power of 135kW, maximum torque of 320 Nm, and its controller is rated for currents up to 500 A.

저속 운전 중 TM의 재생 제동 한계를 실험적으로 보여주기 위해, 400-V 배터리 장치에 연결된 고효율 영구 자석 동기 모터(PMSM)가 발전기로 작동하도록 시험합니다. 

PMSM은 서로 다른 저속에서 작동하도록 하고, 저항 토크, 축 속도 및 배터리 전류가 각 단계에서 기록됩니다. 

PMSM은 0~7700 rpm의 속도 범위에서 작동할 수 있으며, 최대 출력은 135kW, 최대 토크는 320 Nm이며, 그 제어기는 최대 500 A의 전류에 대한 등급을 받았습니다.

 

The results for different operating points of speed from 0 to 1000 rpm with a step size of 25 rpm, and resistive torque from −210 to 0 Nm with a step size of 5 Nm are shown in Fig. 4.

In this figure, a positive current indicates that the battery is recharging while a negative current indicates that the battery is being discharged.

As depicted in Fig. 4, it is observed that despite the fact that the PMSM is operating in the regenerative braking mode of operation, below a specific boundary (red line in Fig. 4), battery current is negative.

This clearly indicates the inefficiency of the regenerative braking process under this boundary.

Hence, by identifying this low-speed boundary and operating above this boundary, energy loss can be minimized during low-speed regenerative braking instances.

속도가 0에서 1000 rpm까지 25 rpm 간격으로, 저항 토크가 -210에서 0 Nm까지 5 Nm 간격으로 다른 작동 점에 대한 결과가 그림 4에 표시되어 있습니다. 

이 그림에서 양의 전류는 배터리가 충전되고 있다는 것을 나타내고, 음의 전류는 배터리가 방전되고 있다는 것을 나타냅니다. 

그림 4에서 나타나듯이, PMSM이 재생 제동 모드에서 작동하더라도 특정 경계 (그림 4의 빨간 선) 아래에서 배터리 전류가 음수임을 관찰할 수 있습니다. 

이는 명백히 이 경계 아래에서 재생 제동 과정의 비효율성을 나타냅니다. 

따라서 이 저속 경계를 식별하고 이 경계 위에서 작동함으로써 저속 재생 제동 상황에서의 에너지 손실을 최소화할 수 있습니다.

 

(4) Proposed EV Brake Controller Considering Low-Speed Boundary

The experimental results obtained from Section III reveal that the TM performance map can serve as a suitable indicator for identifying the dynamically changing low-speed regenerative braking boundary.

Therefore, by taking advantage of this and designing a brake controller in which the low-speed boundary is known for every operating point, energy loss during the braking process can be minimized.

A flowchart representation of the proposed brake controller for a typical front-wheel drive EV is depicted in Fig. 5.

The proposed brake controller takes advantage of the TMregenerative braking performance map as an indicator to adjust the shares of regenerative and friction-based brakes in real-time with the objective of maximizing harvested energy during braking.

제 III 장에서 얻은 실험 결과는 TM 성능 맵이 동적으로 변하는 저속 재생 제동 경계를 식별하는 데 적절한 지표로 활용될 수 있다는 것을 보여줍니다.
따라서 이를 활용하여 저속 경계가 각 작동 점마다 알려진 브레이크 컨트롤러를 설계함으로써 브레이킹 과정 중의 에너지 손실을 최소화할 수 있습니다.
전형적인 전륜 구동 EV를 위한 제안된 브레이크 컨트롤러의 플로우차트 표현이 그림 5에 나와 있습니다.
제안된 브레이크 컨트롤러는 브레이킹 중에 수확된 에너지를 극대화하는 목표로 실시간으로 재생 및 마찰 기반 브레이크의 비율을 조정하기 위해 TM 재생 제동 성능 맵을 지표로 활용합니다.

 

[15]에서 가져온 Ideal 곡선

Based on Fig. 5, the requiredTMtorque (TTM) is used to specify the brake instance such that a negative TM torque (TTM<0) indicates braking.

If brake is requested (TTM<0), front axle (Tbf ) and rear axle (Tbr) brake torques are calculated using the ideal brake torque distribution curve referred to as the I curve [15] to ensure stability and controllability while braking [22].

그림 5를 기반으로 하면, 필요한 TM 토크(TTM)는 브레이크 인스턴스를 지정하는 데 사용됩니다. 

여기서 음의 TM 토크(TTM < 0)는 브레이킹을 나타냅니다. 

브레이크가 요청된 경우 (TTM < 0), 전륜 (Tbf) 및 후륜 (Tbr) 브레이크 토크는 안정성과 조종성을 보장하기 위해 이상적인 브레이크 토크 분배 곡선인 I 곡선 [15]을 사용하여 계산됩니다. 

Considering that the EV under study in this paper is assumed to be front-wheel drive, the front axle brake share (Tbf ) is extracted as an input to the TM regenerative braking performance map.

Based on the output of the map, as long as the operating point of the motor is above the low-speed boundary (Ibat >0) and the motor is capable of meeting the share of front axle brake torque (Tbf ≤ Tregen(max)), the front axle brake share will consist of only regenerative braking (Tregen = Tbf ).

If the required braking torque by the front axle exceeds the motor regenerative braking capability (Tbf > Tregen(max)), regenerative braking is limited to the maximum capacity of the motor (Tregen = Tregen(max)) and the remaining brake torque on the front axle is met by the front axle friction braking.

Once the desired operating point falls belowthe low-speed boundary (Ibat ≤0), regenerative braking share is reduced to zero with a rate of ΔT, and friction braking is increased with the same rate to guarantee a smooth transition between friction and regenerative braking.

본 논문에서 연구 대상인 EV가 전륜 구동이라고 가정하므로 전륜 브레이크 토크(Tbf)는 TM 재생 제동 성능 맵의 입력으로 사용됩니다. 

맵의 출력을 기반으로, 모터의 작동 점이 저속 경계 (Ibat > 0) 위에 있고 모터가 전륜 브레이크 토크(Tbf ≤ Tregen(max))를 충족할 수 있는 경우, 전륜 브레이크 토크는 재생 제동만으로 이루어집니다 (Tregen = Tbf). 

전륜의 필요한 브레이크 토크가 모터의 재생 제동 능력을 초과하는 경우 (Tbf > Tregen(max)), 재생 제동은 모터의 최대 용량으로 제한됩니다 (Tregen = Tregen(max)) 그리고 나머지 브레이크 토크는 전륜 마찰 제동으로 충족됩니다. 

원하는 작동 점이 저속 경계 아래로 떨어지면 (Ibat ≤0), 재생 제동 점유율이 ΔT의 비율로 감소하고 마찰 제동이 동일한 비율로 증가하여 마찰과 재생 제동 사이의 부드러운 전환을 보장합니다. 

With the proposed approach and relying on the TM regenerative braking performance map, optimal brake distribution between regenerative and friction-based brakes can be accomplished.

This method is advantageous due to the fact that the motor performance map can be formed offline with accurate precision during the development stage and using current sensorswith very high resolution and then be implemented in the brake controller, which in turn reduces the computational effort of the brake controller [34].

Furthermore, the proposed approach is based on vehicle speed measurement, which is measured with high accuracywithin the vehicle, and the torque request by the driver, which can be accurately calculated based on the brake pedal displacement.

Therefore, the proposed strategy is easy to implement, accurate, and does not lead to any delays, which significantly speeds up the brake controller performance in yielding optimum brake allocation.

제안된 접근 방식과 TM 재생 제동 성능 맵에 의존하여 재생 및 마찰 기반 브레이크 간의 최적의 브레이크 분배가 이루어집니다. 

이 방법은 개발 단계에서 매우 정확하게 오프라인으로 모터 성능 맵을 형성하고 매우 높은 해상도의 전류 센서를 사용하여 브레이크 컨트롤러의 계산 노력을 줄일 수 있습니다. 

그러므로, 제안된 접근 방식은 차량 내에서 매우 정확하게 측정되는 차량 속도 측정 및 운전자에 의한 토크 요청에 기반하므로 구현이 쉽고 정확하며 지연을 유발하지 않으며 이로 인해 최적의 브레이크 할당을 위해 브레이크 컨트롤러의 성능이 크게 향상됩니다.

 

(5) Experimental Test Bench, Case Studies, And Results

To experimentally validate the effectiveness of the proposed strategy on improving the extracted energy through regenerative braking process, the proposed method is integrated into the brake controller of an HIL experimental test bench.

The experimental setup is shown in Fig. 6 and is represented schematically in Fig. 7.

The test bench consists of two high-performance PMSMs connected to a common shaft with their corresponding inverters.

The PMSMs are the same motors used in Section III to obtain the performance map.

A 400-V battery pack, capable of bidirectional energy flow, is used as the main energy source of this test bench.

Each PMSM and its corresponding inverter is equipped with a cooling loop to maintain the operational temperature within allowable limits.

In this test bench, one of the PMSMs emulates the TM of an EV, while the other one mimics the dynamometer (DYNO) to simulate all the forces imposed on the vehicle including friction braking.

Dyno
: Dynamometer의 줄임말로, 차량의 엔진 또는 동력원의 출력을 측정하는 장비를 가리킵니다. 다이노는 엔진 테스트나 조정, 성능 향상 등의 목적으로 사용됩니다. 엔진 다이노는 엔진을 단독으로 테스트하고 출력을 측정하는 반면, 차량 다이노는 전체 차량을 테스트하여 엔진 출력뿐만 아니라 변속기, 바퀴 및 제동 시스템 등의 효율성을 평가합니다.  

 

Real-time communication is accomplished using the controller area network (CAN) bus.

LabVIEW real-time controller is used as the software for calculating the speed and torque commands at each  set point.

To achieve synchronous operation of both motors, the TMandDYNOoperating points are updated at each step through synchronous commands sent from LabVIEW using CAN.

To experimentally verify the effectiveness of the proposed controller on maximizing the harvested energy during braking, and to delineate its robustness against variability of vehicle/road operating conditions, two different case studies are considered in this section.

제안된 전략이 재생 제동 프로세스를 통해 추출된 에너지를 향상시키는 데 얼마나 효과적인지 실험적으로 검증하기 위해 제안된 방법이 HIL 실험 테스트 벤치의 브레이크 컨트롤러에 통합됩니다. 

실험 설정은 그림 6에 나와 있으며 도식화된 것은 그림 7에 나와 있습니다. 

실험 벤치는 두 개의 고성능 PMSM이 해당 인버터와 공통의 축에 연결된 것으로 구성됩니다. 

PMSM은 섹션 III에서 성능 맵을 얻기 위해 사용된 동일한 모터입니다. 

양방향 에너지 흐름이 가능한 400-V 배터리 팩이 이 실험 벤치의 주요 에너지 원으로 사용됩니다. 

각 PMSM 및 해당 인버터에는 운영 온도를 허용 한도 내로 유지하기 위한 냉각 루프가 장착되어 있습니다. 

이 실험 벤치에서 한 PMSM은 EV의 TM을 흉내 내고, 다른 하나는 마찰 제동을 포함한 차량에 가해지는 모든 힘을 시뮬레이션하기 위해 다이너모(DYNO)를 흉내 냅니다. 

실시간 통신은 컨트롤러 지역 네트워크 (CAN) 버스를 통해 수행됩니다. 

LabVIEW 실시간 컨트롤러는 각 설정 지점에서 속도 및 토크 명령을 계산하는 데 사용됩니다. 

두 모터의 동기화된 작동을 달성하기 위해 LabVIEW를 통해 CAN을 사용하여 각 단계에서 TM 및 DYNO 작동 지점이 동기화된 명령으로 업데이트됩니다. 

브레이크 중에 수확된 에너지를 최대화하기 위한 제안된 컨트롤러의 효과를 실험적으로 검증하고 차량/도로 운영 조건의 변동성에 대한 그 견고성을 명확히하기 위해 이 섹션에서 두 가지 다른 케이스 연구가 고려됩니다.

 

A. Case Study 1: Performance of the Proposed Method

The first case study is conducted to show the effectiveness of the proposed method from the viewpoint of energy savings while comparing the results to a constant low-speed boundary as well as a strategy that utilizes a weight factor to implement variable boundary [31].

This case study emulates a typical EV with a total weight of 1000 kg for three different scenarios.

The first scenario assumes a constant low-speed boundary of 350 rpm, in which the regenerative braking capability is disabled below this speed and all braking is achieved through friction braking.

It is worth mentioning that the fixed low-speed boundary was chosen based on tests carried out formultiple speeds that resulted in the best energy extraction performance for a fixed speed of 350 rpm.

제일 첫 번째 케이스 스터디는 제안된 방법의 효과를 에너지 절약의 관점에서 보여주며 결과를 일정한 저속 경계와 가변 경계를 구현하는 가중치 요소를 사용하는 전략과 비교합니다 [31]. 

이 케이스 스터디는 총 중량이 1000 kg인 전형적인 EV를 세 가지 다른 시나리오로 에뮬레이션합니다. 

첫 번째 시나리오는 350 rpm의 고정된 저속 경계를 가정하며, 이 속도 이하에서는 재생 제동 기능이 비활성화되고 모든 제동이 마찰 제동으로 달성됩니다. 

고정된 저속 경계는 여러 속도에서 수행된 테스트를 바탕으로 선택되었으며, 그 중에서도 350 rpm의 고정된 속도에서 최적의 에너지 추출 성능을 나타냈습니다. 

The second scenario implements the brake force distribution by considering a weight factor that defines the ratio between regenerative and friction braking shares such that a weight factor of one represents a vehicle that all braking on the driven axle is achieved by regenerative braking.

In this approach, depicted in Fig. 8, the weight factor decreases as the vehicle speed is reduced below a predetermined limit of Vmax.

In this study, Vmax is considered to be 30 mph.

두 번째 시나리오는 가변 경계를 구현하기 위해 가중치 요소를 고려하여 제동력 분배를 실시합니다. 여기서 가중치 요소는 구동 축에서의 모든 제동이 재생 제동에 의해 달성되는 차량을 나타내므로 가중치 요소가 하나인 경우입니다.

이 접근 방식은 그림 8에 나와 있으며, 차량 속도가 미리 정해진 Vmax의 제한 아래로 감소함에 따라 가중치 요소가 감소합니다.

이 연구에서 Vmax는 30 mph로 간주됩니다.

The third scenario is carried out by implementing the proposed brake controller.

To provide a fair comparison, all tests are conducted using the Urban Dynamometer Driving Schedule (UDDS) profile.

세 번째 시나리오는 제안된 브레이크 컨트롤러를 구현하여 실시합니다.

공정한 비교를 제공하기 위해 모든 테스트는 Urban Dynamometer Driving Schedule (UDDS) 프로파일을 사용하여 수행됩니다.

 

The amount of energy extracted through regenerative braking process as well as the net energy consumption for completing one UDDS profile is calculated and compared for all scenarios and the results are summarized in Table I.

Based on Table I, the extracted energy through regenerative braking for one UDDS  profile is increased from 223.7 Wh when using constant lowspeed boundary to 228 Wh and 241.8 Wh when utilizing the weight factor brake controller and the proposed braking method, respectively.

As a result, the net energy consumption, which is the difference between the required energy to propel the vehicle and the harvested energy through regenerative braking, is decreased from 1051.3Wh in scenario one to 1047 Wh and 1033.2 Wh for scenarios two and three, respectively.

Based on these results, the proposed brake controller offers an energy saving improvement of 8.09% over the constant low-speed boundary and 6.05% over the weight factor control strategy.

테이블 I에 따르면, UDDS 프로필 하나를 완료하는 데 필요한 에너지 소비량과 함께 재생 제동 과정을 통해 추출된 에너지량을 계산하여 모든 시나리오에 대해 비교합니다. 

테이블 I에 따르면, 상수 저속 경계를 사용할 때 재생 제동을 통해 추출된 에너지량은 223.7 Wh에서 가중치 요소 브레이크 컨트롤러 및 제안된 브레이크 방법을 사용할 때 각각 228 Wh 및 241.8 Wh로 증가합니다. 

결과적으로, 차량을 구동하는 데 필요한 에너지와 재생 제동을 통해 수집된 에너지 간의 차이인 순 에너지 소비는 시나리오 1에서 1051.3Wh에서 시나리오 2와 시나리오 3에서 각각 1047Wh 및 1033.2Wh로 감소합니다. 

이러한 결과에 따르면, 제안된 브레이크 컨트롤러는 상수 저속 경계보다 8.09%의 에너지 절약 향상을 제공하고 가중치 요소 제어 전략보다 6.05%의 에너지 절약을 제공합니다.

 

To visually compare the results of all scenarios, the harvested energy through regenerative braking process is calculated from the experimental results and is illustrated in Fig. 9

Based on these results, it is clear that the gain in the harvested energy is more significant for the proposed method, increasing the divergence as more regenerative braking opportunities become available throughout the driving cycle.

모든 시나리오의 결과를 시각적으로 비교하기 위해 실험 결과에서 재생 제동 과정을 통해 수집된 에너지가 계산되어 그래프 9에 나타냅니다. 

이러한 결과에 따르면, 제안된 방법에 대한 수확된 에너지의 증가는 주행 주기 동안 더 많은 재생 제동 기회가 제공될수록 더욱 크게 나타납니다.

 

Moreover, Figs. 10-12 illustrate the experimental results of this study for the duration of 20 s of the UDDS profile for each scenario.

The 20-s sample time, during which the vehicle is decelerating from 34.6 mph to a complete stop, was selected in order to clearly show the difference between the results of all scenarios.

Based on these plots, it is evident that in all tests, the total amount of brake torque is the same.

However, the regenerative braking torque share in scenario three is higher than that of scenarios one and two.

As expected, this is due to the fact that the proposed strategy implemented in scenario three achieves blending at an optimum point whereas in scenario one this point is only dependent on a fixed speed and in scenario two this point is only proportional to the vehicle’s speed.

As a result, more energy is harvested in scenario three during braking instances.

또한, 그림 10-12는 각 시나리오의 UDDS 프로파일의 20초 동안의 실험 결과를 보여줍니다. 

차량이 34.6 mph에서 완전히 정지할 때까지 감속하는 20초 샘플 시간은 모든 시나리오의 결과를 명확하게 비교하기 위해 선택되었습니다. 

이 그림을 통해 모든 테스트에서 총 제동 토크의 양이 동일함을 알 수 있습니다. 

그러나 시나리오 세 개 중 시나리오 삼의 재생 제동 토크 비율이 시나리오 일과 이보다 높은 것을 확인할 수 있습니다. 

이는 시나리오 삼에서 제안된 전략이 최적 지점에서 혼합을 달성하는 반면, 시나리오 일에서 이 지점이 고정 속도에만 의존하고 시나리오 이에서는 이 지점이 차량의 속도에만 비례하기 때문으로 예상됩니다. 

결과적으로, 시나리오 삼에서는 제동 시 에너지가 더 많이 회수됩니다. 

 

This energy harvesting improvement is more notable when considering a typical annual driving range of 12,000 miles for an EV [15].

Assuming that the EV under study travels this distance solely based on UDDS profile, by taking advantage of the proposed brake controller, an extra 29.15 kWh of energy can be recovered in one year compared to a similar case of considering constant low-speed boundary.

This energy recovery in one year is adequate to meet the EV energy requirements for an extra 28.22 UDDS profiles, which is equivalent to driving 210.24 more miles.

이 에너지 회수 개선은 전기 자동차의 전형적인 연간 주행 거리인 12,000마일을 고려할 때 더욱 두드러집니다. 

이 연구 대상의 전기 자동차가 이 거리를 UDDS 프로파일만으로 달성한다고 가정하면, 제안된 브레이크 컨트롤러를 활용하여 연간 29.15 kWh의 추가 에너지를 회수할 수 있으며, 이는 고정된 저속 경계를 고려하는 유사한 경우보다 1년 동안 더 많은 에너지를 회수할 수 있음을 의미합니다. 

이 1년 동안의 에너지 회수는 전기 자동차의 에너지 요구를 추가로 충족시키기에 충분하며, 이는 210.24마일을 운전하는 것과 동등합니다.

 

B. Case Study 2: Robustness of the Proposed Method

Vehicle/road operating conditions can vary depending on many factors including the number of passengers or cargo that the vehicle is carrying.

In order to show the robustness of the proposed method against changes in vehicle conditions, two different scenarios are experimentally tested.

In the first scenario, the total vehicle weight is considered to be 1000 kg and in the second scenario, the weight is increased to 1500 kg.

In both scenarios, the proposed brake control strategy is implemented, and the UDDS profile is utilized as the reference driving profile.

차량/도로 운영 조건은 차량이 운반하는 승객이나 화물의 수 등 많은 요소에 따라 달라질 수 있습니다. 

제안된 방법이 차량 조건의 변화에 대한 강건성을 보이기 위해 두 가지 다른 시나리오가 실험적으로 테스트되었습니다. 

첫 번째 시나리오에서는 총 차량 무게가 1000 kg로 고려되고, 두 번째 시나리오에서는 무게가 1500 kg로 증가합니다. 

두 시나리오 모두에서 제안된 브레이크 제어 전략이 시행되며, 참조 주행 프로파일로는 UDDS 프로파일이 사용됩니다. 

 

Fig. 13 shows the UDDS profile with the obtained low-speed boundary points calculated by the brake controller.

As depicted in this figure, the low-speed boundary points vary under different vehicle weights and thus the proposed method is robust against changes in vehicle’s weight.

From the results of Fig. 13, it is also observed that the low-speed boundary for scenario two occurs at higher speeds relative to scenario one, which not only confirms the fact that low-speed regenerative braking boundary changes under different conditions, but also indicates that the effective range of regenerative braking is reduced as the vehicle weight increases.

그림 13은 브레이크 컨트롤러에 의해 계산된 저속 경계 점이 표시된 UDDS 프로파일을 보여줍니다. 

이 그림에서 보듯이, 저속 경계 점은 차량 무게가 다를 때 달라지며, 따라서 제안된 방법은 차량 무게의 변화에 강건합니다. 

그림 13의 결과에서 또한 두 번째 시나리오의 저속 경계가 첫 번째 시나리오에 비해 더 높은 속도에서 발생함을 관찰할 수 있는데, 이는 저속의 재생 제동 경계가 다른 조건 하에서 변화한다는 사실을 확인할 뿐만 아니라 차량 무게가 증가함에 따라 재생 제동의 효과적인 범위가 축소된다는 것을 나타냅니다.

 

For this case study, the amount of energy extracted through regenerative braking process as well as the net energy consumption for completing one UDDS is calculated for both scenarios and the results are summarized in Table II.

Despite the fact that the effective range of regenerative braking is reduced for an increased vehicle weight, the harvested energy through regenerative braking process is increased from 241.8 Wh to 266.8 Wh as the vehicle weight increases.

This is due to the fact that a heavier vehicle requires a higher regenerative braking torque for deceleration and therefore more energy can be extracted during braking.

Similarly, since the required energy for propelling the vehicle in scenario two is more than that of scenario one, the net energy consumption is also increased from 1033.2Wh to 1994.2 Wh.

이 경우 연구에서는 재생 제동 과정을 통해 추출된 에너지 양과 UDDS를 완료하기 위한 순 에너지 소비량을 두 시나리오 모두에 대해 계산하고 결과를 테이블 II에 요약했습니다. 

차량 무게가 증가함에 따라 재생 제동의 효과적인 범위가 축소되었음에도 불구하고, 재생 제동 과정을 통해 추출된 에너지는 차량 무게가 증가함에 따라 241.8Wh에서 266.8Wh로 증가했습니다. 

이는 보다 무거운 차량은 감속을 위해 더 높은 재생 제동 토크가 필요하며, 따라서 제동 중에 더 많은 에너지를 추출할 수 있기 때문입니다. 

마찬가지로, 시나리오 2의 차량을 운행하는 데 필요한 에너지가 시나리오 1보다 많기 때문에 순 에너지 소비량도 1033.2Wh에서 1994.2Wh로 증가합니다.

 

(6) Conclusion

In this paper, the concept behind low-speed regenerative braking limitation of EVs was mathematically investigated and the influence of considering a low-speed boundary during braking was analyzed from the viewpoint of energy extraction.

A brake controller was proposed that takes advantage of the motor performance map to determine the dynamic low-speed regenerative braking boundary.

The effectiveness of the proposed control strategy on maximizing the amount of extracted energy through regenerative braking process was verified using an experimental HIL motor/dynamometer test bench.

The results revealed that using the proposed method, the amount of harvested energy increased compared to a case where a constant low-speed boundary was taken into account and a case where the low-speed boundary was calculated through a variable weight factor.

The robustness of the proposed controller against vehicle operating conditions was also experimentally evaluated by considering two scenarioswith different vehicleweights.

The results showed that the proposed brake controller can accurately detect low-speed boundary points which occurred at higher speeds as the vehicle weight increased.

The results of this study can be beneficial for EV automotive companies and related research institutions looking into improving regenerative braking capability of their existing or future vehicles in order to improve the driving range of EVs.

이 논문에서는 EV의 저속 재생 제동 제한 개념이 수학적으로 조사되었으며, 제동 중 저속 경계를 고려하는 것이 에너지 추출 관점에서 어떻게 영향을 미치는지 분석되었습니다. 

제안된 제동 컨트롤러는 모터 성능 맵을 활용하여 동적 저속 재생 제동 경계를 결정합니다. 

제안된 제어 전략이 재생 제동 과정을 통한 추출된 에너지량을 극대화하는 데 미치는 효과는 실험적인 HIL 모터/다이나모 시험 벤치를 사용하여 확인되었습니다. 

결과는 제안된 방법을 사용하여 추출된 에너지 양이 일정한 저속 경계를 고려한 경우 및 변수 가중치 요소를 사용하여 계산된 경우보다 증가했음을 보여주었습니다. 

제안된 컨트롤러의 견고성은 차량 무게가 다른 두 가지 시나리오를 고려하여 실험적으로 평가되었습니다. 

결과는 제안된 제동 컨트롤러가 차량 무게가 증가함에 따라 발생하는 저속 경계 지점을 정확하게 감지할 수 있음을 보여주었습니다. 

이 연구의 결과는 EV 자동차 회사와 관련 연구 기관이 기존이나 미래의 차량의 주행 거리를 향상시키기 위해 재생 제동 능력을 개선하는 것을 고려하는 데 도움이 될 수 있습니다.