Paper/Electrical and Electrical Engineering

[논문리뷰] Maximizing Regenerative Braking Energy Recovery of Electric Vehicles Through Dynamic Low-Speed Cutoff Point Detection <2>

얼죽아여뜨샤 2024. 4. 7. 17:01

0. 원문

Maximizing_Regenerative_Braking_Energy_Recovery_of_Electric_Vehicles_Through_Dynamic_Low-Speed_Cutoff_Point_Detection.pdf
3.19MB

1. 내용

(0) Abstract

This paper introduces a novel approach for dynamically detecting the lowest speed threshold at which regenerative braking is effective in electric vehicles (EVs).

The control approach is based on real-time sensing of the motor controller dc link current and disabling regenerative braking when current changes direction while the motor is operating as a generator.

Various factors influencing the regenerative braking capability of EVs at low speed are discussed, and the simulation studies are carried out to illustrate the effect of each factor on the displacement of the low-speed threshold.

Based on the results obtained from the simulation studies, a dynamic low-speed cutoff point (LSCP) detection method is proposed.

This method requires no hardware modification to the vehicle braking architecture and can be implemented solely by modifying the brake controller.

The proposed method is tested on an experimental EV test platform for a predetermined drive cycle.

It is shown that in comparison to considering a constant low-speed threshold during braking, the amount of energy recaptured through the regenerative braking process can be improved by taking advantage of the proposed method.

이 논문은 전기 자동차(EVs)에서 재생 제동이 효과적인 최저 속도 임계값을 동적으로 감지하는 새로운 방법을 제안합니다. 

이 제어 방법은 모터 컨트롤러 DC 링크 전류의 실시간 감지를 기반으로하며, 모터가 발전기로 작동하는 동안 전류가 방향을 변경할 때 재생 제동을 비활성화합니다. 

저속에서 EV의 재생 제동 능력에 영향을 주는 다양한 요인에 대해 논의되고, 각 요인이 저속 임계값의 이동에 미치는 영향을 설명하기 위해 시뮬레이션 연구가 수행됩니다. 

시뮬레이션 연구에서 얻은 결과를 기반으로 동적 저속 절단점(LSCP) 감지 방법이 제안됩니다. 

이 방법은 차량 제동 구조에 하드웨어 수정이 필요하지 않으며, 제동 컨트롤러를 수정함으로써 단독으로 구현할 수 있습니다. 

제안된 방법은 미리 정의된 주행 주기에 대해 실험적 EV 테스트 플랫폼에서 테스트되었습니다. 

제안된 방법을 활용하면 제동 중에 일정한 저속 임계값을 고려하는 것과 비교하여 재생 제동 프로세스를 통해 회수되는 에너지 양을 향상시킬 수 있음을 보여줍니다.

 

#Keyword : Brake controller, electric vehicle (EV), friction braking, recaptured energy, regenerative braking, test bench.

 

(1) Introduction

Recently, the automotive companies are confronting the challenge of reducing fuel usage and emissions caused by the transportation sector.

In fact, the transportation sector alone accounts for roughly 21% of the total energyrelated emissions [1].

To accomplish this, vehicle electrification is considered to be the most viable solution [2].

Vehicle electrification started its evolution from 1950 and has influenced major parts of the vehicle system including electric power steering, brakes, and powertrain [3].

In recent years, growing sales of electric vehicles (EVs) have added momentum to the global shift toward transportation electrification, making the automotive industry invest more heavily in the transportation electrification technology [4].

Nevertheless, one important obstacle in the large-scale adoption of EVs is their restricted driving range [5].

Hence, extensive research in both industry and academia is carried out with the aim of improving the efficiency and driving range of these vehicles [5]–[7].

최근에는 자동차 회사들이 운송 부문에서 발생하는 연료 소비와 배출물을 줄이는 과제에 직면하고 있습니다. 

실제로, 운송 부문은 전체 에너지 관련 배출물의 약 21%를 차지합니다. 

이를 이루기 위해 차량 전기화가 가장 타당한 해결책으로 간주됩니다. 

차량 전기화는 1950년대부터 진화를 시작했으며, 전기식 파워 스티어링, 브레이크 및 파워트레인을 포함한 주요 부분에 영향을 미쳤습니다. 

최근 몇 년간 전기 자동차(EVs)의 판매가 증가함에 따라 전 세계적으로 운송 전기화로의 전환이 가속화되고 있으며, 이로 인해 자동차 산업은 운송 전기화 기술에 대해 더 많은 투자를 하고 있습니다. 

그러나 대규모 EV 채택의 중요한 장애물 중 하나는 그들의 제한된 주행 거리입니다. 

따라서 산업 및 학계에서는 이러한 차량의 효율성과 주행 거리를 향상시키기 위한 광범위한 연구가 이루어지고 있습니다. 

As opposed to conventional vehicles, in which the brake energy is dissipated as heat, the regenerative braking capability of EVs can assist in recapturing and storing this energy during braking [8], [9].

Regenerative braking is defined as recharging the main energy source by feeding back energy from the traction motor (TM) toward the battery during brake instances [8], [10].

It is shown that the driving range of EVs can be extended by 15% through the regenerative braking process [8], and this amount can be even more significant in metropolitan driving conditions, where the brakes are used more frequently [11].

In other words, the regenerative braking application improves the braking performance and extends the driving range without any additional cost [12].

Even though a fully electrified regenerative braking system is feasible for EVs [13], to fulfill simultaneous energy recovery and maintain vehicle stability during the braking instance, both regenerative- and friction-based brakes have to coexist [14]–[18].

Taking into account the typical delay associated with friction-based brakes, which can be problematic, especially in vehicle slip control, simultaneously applying regenerative braking can significantly enhance the braking performance due to faster response time.

Moreover, under heavy braking circumstances, the resistive force produced by the regenerative braking process may not be sufficient to overcome the required force to slow down the vehicle, making friction braking an essential part of the braking system [15].

Furthermore, the friction-based brakes are still vital in maintaining vehicle safety in case of possible failure of the electrical system [19].

일반적인 차량과는 달리, EV의 재생 제동 능력은 브레이크 에너지가 열로 분산되는 대신 제동 중에 이 에너지를 회수하고 저장하는 데 도움이 될 수 있습니다. 

재생 제동은 브레이크 순간에 추진 모터(TM)에서 배터리로 에너지를 되돌려주는 것으로 정의됩니다. 

재생 제동 과정을 통해 EV의 주행 거리를 15% 연장할 수 있으며, 이런 양상은 특히 브레이크가 더 자주 사용되는 대도시 주행 조건에서 더 중요합니다. 

다시 말해, 재생 제동 적용은 추가 비용없이 브레이킹 성능을 향상시키고 주행 거리를 연장합니다. 

비록 EV에 완전히 전기화된 재생 제동 시스템이 가능하다고 해도, 브레이킹 순간에 동시에 에너지 회수와 차량 안정성을 유지하기 위해 재생 및 마찰 기반 브레이크가 공존해야 합니다. 

특히 차량 슬립 제어에서 문제가 될 수 있는 마찰 기반 브레이크의 전형적인 지연을 고려하면, 재생 제동을 동시에 적용함으로써 브레이킹 성능이 크게 향상될 수 있습니다. 

또한, 심한 브레이킹 상황에서 재생 제동 과정에서 생성되는 저항력이 차량을 감속하기 위해 필요한 힘을 극복할 만큼 충분하지 않을 수 있으므로, 마찰 브레이크는 브레이킹 시스템의 중요한 부분입니다. 

더 나아가, 전기 시스템의 가능한 고장을 고려할 때 마찰 기반 브레이크는 여전히 차량 안전을 유지하는 데 중요합니다.

 

Different methods have been proposed in the literature to maximize the regenerative braking capability of EVs and hybrid EVs (HEVs) [20], [21].

In [20], the electric motor’s operating point is adjusted by a downshifting system, which results in maximizing regenerative braking.

Downshifting system
: 다운시프트(downshift)는 커브길을 가거나 언덕길을 올라갈 때 자동차의 변속기어를 고단에서 저단으로 바꾸어 속도를 줄이는 것을 뜻한다.

 

In [21], regenerative braking is improved in an HEV by considering a battery aging model.

Regarding the maximum energy that can be harvested through the regenerative braking process, one important aspect is the inability to recharge the energy source at very low speeds.

Although functioning in regenerative braking mode at low speeds can still be achieved by operating the TM as a generator, the motor and brake controller are no longer capable of efficiently transferring this energy to the battery.

Thus, the current is extracted from the battery instead of charging it, and this, in turn, results in energy loss [10], [15].

As a result, the regenerative braking is considered inefficient during these instances and braking should solely be realized by friction brakes [14], [15].

Moreover, this low-speed threshold, under which regenerative braking is no longer effective, depends on a variety of factors and varies under different operating conditions [8], [15].

Therefore, a dynamically changing low-speed threshold should be considered in designing the brake controller [15].

문헌에서는 EV와 하이브리드 EV(HEV)의 재생 제동 능력을 극대화하기 위한 다양한 방법이 제안되었습니다. 

[20]에서는 전기 모터의 운전 점이 다운시프팅 시스템에 의해 조정되어 재생 제동을 극대화합니다. 

[21]에서는 HEV에서 배터리 노화 모델을 고려하여 재생 제동이 개선됩니다. 

재생 제동 과정을 통해 회수할 수 있는 최대 에너지에 관한 중요한 측면 중 하나는 매우 낮은 속도에서 에너지원을 충전할 수 없는 것입니다. 

낮은 속도에서 재생 제동 모드로 작동하는 것은 여전히 TM을 발전기로 작동시킬 수 있지만, 모터 및 브레이크 컨트롤러는 이 에너지를 배터리로 효율적으로 이전할 수 없게 됩니다. 

따라서 배터리에서 전류가 추출되어 충전되지 않으며, 결과적으로 에너지가 손실됩니다. 

이로 인해 이러한 경우에 재생 제동이 비효율적으로 간주되며, 브레이킹은 마찰 브레이크로만 실현되어야 합니다. 

또한, 재생 제동이 더 이상 효과적이지 않은 낮은 속도 임계값은 다양한 요인에 따라 변하며 다른 운전 조건에서도 변할 수 있습니다. 

따라서 동적으로 변경되는 낮은 속도 임계값을 브레이크 컨트롤러 설계에 고려해야 합니다. 

The effect of low-speed limitation in blending regenerative- and friction-based brakes has been comprehensively examined and appraised in the literature.

In [8], a control strategy is applied to the EV brake force distribution between regenerative- and friction-based brakes at low speeds that is based on fuzzy logic control.

In [22] and [23], considering that only a small portion of the total energy can be saved at low speeds, a fixed speed is considered as a low-speed threshold.

Naseri et al. [18] have used an artificial neural network control method to reduce the regenerative braking torque at low speeds.

A weight factor has been used in [24] to reduce the proportion of regenerative-based brakes in comparison to friction-based brakes at low speeds.

In all of the aforementioned studies, the dynamically changing nature of the low-speed threshold is overlooked.

재생 및 마찰 기반 브레이크의 혼합에 대한 낮은 속도 제한의 영향이 종합적으로 검토되고 평가되었습니다. 

[8]에서는 낮은 속도에서 EV 브레이크 힘 분배에 퍼지 논리 제어를 기반으로 한 제어 전략이 적용됩니다. 

[22]와 [23]에서는 낮은 속도를 낮은 속도 임계값으로 고려하여 전체 에너지의 작은 부분만 절약될 수 있음을 고려합니다. 

Naseri et al. [18]은 인공 신경망 제어 방법을 사용하여 낮은 속도에서 재생 제동 토크를 감소시켰습니다. 

[24]에서는 낮은 속도에서 마찰 기반 브레이크 대비 재생 기반 브레이크의 비율을 줄이기 위해 가중치 요소가 사용되었습니다. 

앞서 언급한 모든 연구에서는 낮은 속도 임계값의 동적인 변경 특성을 고려하지 않았습니다. 

This paper examines different factors affecting a change in low-speed regenerative braking performance of EVs and proposes a novel method that can accurately identify the instance where regenerative braking is no longer effective.

The proposed approach is validated using a hardware-in-the-loop (HIL) EV test bench setup designed to emulate the EV performance.

Therefore, the main technical contributions of this paper are: 1) to outline the dynamic nature of the low-speed regenerative braking threshold and show that this boundary can change under different circumstances and 2) to propose a novel method based on the dc link current monitoring that dynamically determines the low-speed boundary and can improve energy harvesting during regenerative braking at low speeds.

본 논문은 EV의 낮은 속도 재생 제동 성능이 변경되는데 영향을 미치는 다양한 요소를 검토하고, 재생 제동이 더 이상 효과적이지 않은 시점을 정확하게 식별할 수 있는 새로운 방법을 제안합니다.

제안된 접근 방식은 EV 성능을 에뮬레이트하기 위해 설계된 하드웨어 인 더 루프(HIL) EV 테스트 벤치 세트업을 사용하여 검증됩니다.

따라서 본 논문의 주요 기술적 기여는 다음과 같습니다: 1) 낮은 속도 재생 제동 임계값의 동적 성격을 개요화하고, 이 경계가 다양한 상황에서 변할 수 있음을 보여주며, 2) dc 링크 전류 모니터링을 기반으로 낮은 속도 경계를 동적으로 결정하고, 낮은 속도에서 재생 제동 중 에너지 회수를 향상시킬 수 있는 새로운 방법을 제안합니다.

 

The remainder of this paper is organized as follows. In Section II, the EV brake force distribution and the restrictions associated with the regenerative braking capability at low speeds are explained.

Section III discusses various factors influencing low-speed regenerative braking performance and investigates the effect of each factor on regenerative braking energy harvesting capability using an EV simulation model.

The proposed EV brake controller is explained in detail in Section IV.

In Section V, the HIL experimental test platform for evaluating the performance of the proposed EV brake controller is explained.

Experimental results are presented in Section VI followed by a discussion on the advantages of utilizing the proposed brake strategy from the viewpoint of energy savings.

Finally, in Section VII, conclusions are drawn.

논문의 나머지 부분은 다음과 같이 구성되어 있습니다. 

제2장에서는 EV 브레이크 힘 분배와 낮은 속도에서의 재생 제동 능력과 관련된 제한 사항에 대해 설명합니다. 

제3장에서는 다양한 요인이 낮은 속도 재생 제동 성능에 미치는 영향을 논의하고 EV 시뮬레이션 모델을 사용하여 각 요인이 재생 제동 에너지 회수 능력에 미치는 영향을 조사합니다. 

제안된 EV 브레이크 컨트롤러에 대한 자세한 설명은 제4장에 있습니다. 

제5장에서는 제안된 EV 브레이크 컨트롤러의 성능을 평가하기 위한 HIL 실험 플랫폼에 대해 설명합니다. 

제6장에서는 실험 결과를 제시하고, 제안된 브레이크 전략의 에너지 절약 측면에서의 장점에 대한 토론을 제공합니다. 

마지막으로, 제7장에서 결론을 도출합니다.

 

(2) EV Brake Force Distribution and Low-Speed Limitation

Most EVs and HEVs exert both an ordinary (friction-based) and regenerative-based braking system.

The friction-based brakes typically consist of frictional drum or disk braking assemblies actuated by a hydraulic system.

On the other hand, regenerative-based braking utilizes the TM to provide a resistive torque to the driven wheels and recharge the battery by converting the kinetic energy of the vehicle to electricity.

Friction drum
: 자동차 또는 차량의 제동 시스템에서 사용되는 장비입니다. 이 장비는 제동 힘을 시뮬레이션하여 제동 시스템의 성능을 테스트하고 평가하는 데 사용됩니다. 일반적으로 제동 힘이 바퀴에 작용하는 과정을 모방하기 위해 회전하는 원통 모양의 드럼 형태로 디자인되어 있습니다.
Disk braking assemblie
: 차량의 제동 시스템 중 하나로, 회전하는 디스크(브레이크 로터)를 사용하여 차량을 제동하는 장비를 가리킵니다. 디스크 브레이크 어셈블리는 일반적으로 브레이크 캘리퍼, 브레이크 패드, 브레이크 로터 등의 구성 요소로 구성됩니다.

 

The dissipation of kinetic energy during braking instances can be advantageously recuperated by controlling the power electronics used in the TM (Transmission) controller such that the TM operates as a generator [25].

The goal of simultaneous utilization of both regenerative- and friction-based brakes is to accomplish energy recovery and maintain vehicle stability during braking instance [14], [15].

Moreover, to achieve the shortest braking distance, while maintaining the vehicle stability, precise allocation of the entire braking force between front and rear axles is crucial.

Inaccurate distribution of this force between the front and rear axle may result in locking of either the front or rear wheels due to excessive brake force and lead to instability [26].

Thus, designing a brake controller in which accurate distribution between regenerative braking and friction braking as well as between the front and rear axles are considered is of extreme importance.

대부분의 전기차(EVs)와 하이브리드 전기차(HEVs)는 보통의(마찰 기반) 및 재생 기반 브레이킹 시스템을 모두 사용합니다. 

마찰 기반 브레이크는 일반적으로 유압 시스템에 의해 작동되는 마찰 드럼이나 디스크 브레이킹 어셈블리로 구성됩니다. 

반면, 재생 기반 브레이킹은 주행 바퀴에 저항 토크를 제공하고 차량의 운동 에너지를 전기로 변환하여 배터리를 충전하는데 TM을 사용합니다. 

브레이킹 순간에 운동 에너지의 소멸은 TM(변속기) 컨트롤러에서 전력 전자 기기를 제어하여 TM이 발전기로 작동하도록 하는 방식으로 유리하게 회복될 수 있습니다. 

재생 및 마찰 기반 브레이크를 동시에 사용하는 목표는 브레이킹 순간에 에너지 회수와 차량 안정성을 유지하는 것입니다. 

또한 차량 안정성을 유지하면서 가능한 가장 짧은 브레이킹 거리를 달성하기 위해 전체 브레이킹 힘을 정확하게 전방과 후방 축 사이에 분배하는 것이 중요합니다. 

전방과 후방 축 사이의 이 힘의 정확한 분배는 브레이크 힘의 과도한 분배로 인해 전방 또는 후방 바퀴가 잠겨 불안정성을 초래할 수 있기 때문에 극도로 중요합니다. 

따라서 재생 브레이킹과 마찰 브레이킹 사이의 정확한 분배뿐만 아니라 전방과 후방 축 사이의 분배를 모두 고려하는 브레이크 컨트롤러를 설계하는 것이 매우 중요합니다. 

The latter distribution is calculated by taking into account the load transfer from the rear axle to the front axle during the braking instance, which leads to a normalized nonlinear hyperbolic curve referred to as the I curve, as depicted in Fig. 1 [26].

The former distribution depends on the displacement of the driven axle of an EV.

In other words, only the driving axle share is available for energy extraction through the regenerative braking process [27], [28].

후방 축에서 전방 축으로의 하중 이전을 고려하여 계산되며, 이는 그림 1에 나와 있는 I 곡선이라는 정규화된 비선형 초변곡선을 초래합니다. 

전자 분배는 EV의 구동 축의 변위에 따라 달려 있습니다. 

다시 말해, 재생 브레이킹 프로세스를 통해 에너지를 추출하는 데는 운전 축의 비중만 사용할 수 있습니다. 

It is worth mentioning that since a front-wheel drive EV is considered in this paper, the braking forces imposed on the front axle consist of both regenerative-based and frictionbased brakes, while the rear axle brake force is solely met by friction-based brakes, as shown in Fig. 2.

이 논문에서 전륜 구동 EV가 고려되었으므로 전방 축에 가해지는 브레이크 힘은 재생 및 마찰 기반 브레이크 모두이며, 후방 축 브레이크 힘은 오로지 마찰 기반 브레이크로 충족됩니다.

 

One of the most important design constraints regarding regenerative braking, especially while efficiency and energy consumption are taken into consideration, is the limitation of the EV TM to function as a generator below a certain speed, referred to as low-speed cutoff point (LSCP) [15].

While decelerating, regenerative braking makes use of the back electromotive force (EMF) of the TM as a voltage source to recharge the main energy source.

However, at low speeds, due to a lack of EMF produced by the TM, the regenerative braking process is incapable of carrying out this task.

The low-speed threshold highly relies on the specifications of the TM, energy storage voltage and state of charge level, driver command, and driving conditions.

Below this speed threshold, although the TM is still capable of imposing a resistive torque to decelerate the vehicle, energy cannot be recaptured through the regenerative braking process [10].

재생 브레이킹을 고려할 때 특히 효율성과 에너지 소비가 고려될 때 가장 중요한 설계 제약 중 하나는 저속에서 EV TM이 발전기로 작동하는 것을 제한하는 것입니다. 이를 저속 차단점(LSCP)이라고 합니다. 

감속 중에 재생 브레이킹은 TM의 역기전력(EMF)을 주전력원을 재충전하기 위한 전압원으로 사용합니다. 

그러나 저속에서는 TM이 생산하는 EMF의 부족으로 인해 재생 브레이킹 프로세스가 이 작업을 수행할 수 없습니다. 

저속 임계값은 TM의 사양, 에너지 저장 전압 및 충전 상태 수준, 운전자 명령 및 주행 조건에 크게 의존합니다. 

이 속도 임계값 아래에서는 TM이 여전히 차량을 감속시키기 위해 저항 토크를 가할 수 있지만, 재생 브레이킹 프로세스를 통해 에너지를 회수할 수는 없습니다.

 

Based on Fig. 3, until point A where the shaft speed is almost 200 rpm, the motor controller dc link current is negative, indicating that the battery is being recharged through regenerative braking.

However, after point A, although the motor is still producing the required resistive force for deceleration, the motor controller dc link current changes direction.

This means that the current is no longer pushed back into the battery and instead is extracted from it, which in turn leads to an energy loss equivalent to area B.

Fig. 3에 따르면, 축 속도가 거의 200 rpm이 되는 A 지점까지, 전동기 컨트롤러 DC 링크 전류가 음수이며, 이는 재생 브레이킹을 통해 배터리가 충전되고 있음을 나타냅니다. 

그러나 A 지점 이후에는, 전동기는 여전히 감속을 위해 필요한 저항 토크를 생성하지만, 전동기 컨트롤러 DC 링크 전류가 방향을 바꿉니다. 

이는 전류가 더 이상 배터리로 되돌려지지 않고 대신 배터리에서 추출된다는 것을 의미하며, 결과적으로 B 영역에 해당하는 에너지 손실이 발생합니다.

 

(3) Factors Affecting Regenerative Braking Cufoff Point

Taking into account that the regenerative braking capability of the TM at low speed is influenced by the speed and requested resistive force during deceleration, it is important to analyze the forces acting on a vehicle during this instance. 

According to Newton’s second law of motion, the forces acting on a vehicle can be expressed by the following equation [29]:

TM의 저속 재생 브레이킹 능력은 속도와 감속 중 요청된 저항력에 의해 영향을 받기 때문에, 이 순간 차량에 작용하는 힘을 분석하는 것이 중요합니다. 

뉴턴의 운동 법칙에 따르면, 차량에 작용하는 힘은 다음과 같이 표현될 수 있습니다.

where FD denotes the driving force in N, m is the total vehicle mass in kg, fr is the rolling resistance coefficient, α is the road slope angle, ρα is the mass density of the air in kg/m3, CD is the aerodynamic drag coefficient, which characterizes the shape of EV, A f is the frontal area of the vehicle in m2, v is the vehicle speed in m/s, vW is the wind speed in m/s, and a is the acceleration or deceleration of the vehicle in m/s2 [26].

식 (1)에서 FD는 N으로 표시되며, m은 kg 단위의 총 차량 질량을 나타냅니다. fr은 굴러가는 저항 계수를 나타내며, α는 도로 경사 각도를 나타냅니다. ρα는 공기의 질량 밀도로, kg/m3으로 표시됩니다. CD는 EV의 모양을 특성화하는 공기 저항 계수로, Af는 차량의 전면 면적을 m2로 나타냅니다. v는 차량 속도를 m/s로, vW는 풍속을 m/s로, a는 차량의 가속도 또는 감속도를 m/s2로 나타냅니다 [26]. 

The acting forces on a typical vehicle are depicted in Fig. 4.

According to (1), variables such as vehicle mass (m), road slope angle (α), vehicle speed (v), wind speed (vW), and acceleration/deceleration (a) are the main factors that can have an impact on the net force acting on the vehicle (FD) [30].

전형적인 차량에 작용하는 힘은 그림 4에 나타납니다. 

식 (1)에 따르면, 차량 질량 (m), 도로 경사 각도 (α), 차량 속도 (v), 풍속 (vW), 및 가속도/감속도 (a)와 같은 변수들이 차량에 작용하는 순 힘 (FD)에 영향을 미칠 수 있는 주요 요인입니다 [30].

 

In order to investigate the impact of each factor on the displacement of LSCP, a test bench simulation model is utilized in MATLAB/SIMULINK to emulate the EV operation during braking.

The simulation model is shown in Fig. 5 and consists of a 400-V Li-ion battery model, a control block, and two permanent magnet synchronous motor (PMSM) drives coupled together through a mechanical shaft model.

One motor is controlled to emulate the vehicle TM and the other mimics all the forces acting on the vehicle due to vehicle movement and acceleration/deceleration.

각 요인이 LSCP의 변위에 미치는 영향을 조사하기 위해 MATLAB/SIMULINK에서 시험장 시뮬레이션 모델을 활용합니다. 

시뮬레이션 모델은 그림 5에 나타나며, 400-V 리튬이온 배터리 모델, 제어 블록 및 두 개의 영구 자석 동기 모터 (PMSM) 드라이브로 구성되어 있습니다. 이들은 기계적인 축 모델을 통해 함께 결합됩니다. 

하나의 모터는 차량 TM을 흉내 내기 위해 제어되고, 다른 하나는 차량 이동 및 가속도/감속도로 인한 차량에 작용하는 모든 힘을 모방합니다.

 

In order to simulate a drive cycle, the speed reference of the TM is calculated from the drive cycle speed by taking into account the overall vehicle gear ratio (G).

Moreover, the net imposed force on the EV is calculated based on (1), and after converting to a torque signal it is used as a reference for the Dyno to emulate road conditions.

In this configuration, the TM and Dyno are operated in speed control and torque control modes, respectively.

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.

 

드라이브 사이클을 시뮬레이션하기 위해 TM의 속도 참조값은 전체 차량 기어 비율 (G)을 고려하여 드라이브 사이클 속도에서 계산됩니다. 

또한, EV에 가해지는 순 힘은 (1)식을 기반으로 계산되고, 토크 신호로 변환한 후에는 도로 조건을 모방하기 위해 Dyno에 참조값으로 사용됩니다. 

이 구성에서 TM과 Dyno는 각각 속도 제어 및 토크 제어 모드로 작동됩니다.

 

Fig. 6 shows the representation of the controller used for this simulation.

In the controller algorithm, FD is calculated at every instance as the total required vehicle force acting on the vehicle, and based on its sign, it is determined whether braking is needed or not.

In this case, a negative driving force (FD < 0) represents braking.

During braking, the brake force share of the front and rear axles are calculated from FD using the I curve as discussed in Section II.

Since the vehicle under study is a front-axle drive EV, the front axle share of braking force, which is available to the motor during regenerative braking, is converted to torque and given to the Dyno as the torque reference during braking.

Fig. 6는 이 시뮬레이션에 사용된 컨트롤러의 표현을 보여줍니다. 

컨트롤러 알고리즘에서는 FD가 차량에 작용하는 총 필요한 힘을 각 순간마다 계산하고 그 부호에 따라 브레이크가 필요한지 여부를 결정합니다. 

이 경우, 음의 운전력 (FD < 0)은 브레이킹을 나타냅니다. 

브레이킹 중에는 섹션2에서  I 곡선에서 설명된대로 FD를 사용하여 앞뒤 축의 브레이크 힘의 분배가 계산됩니다. 

이 연구 대상 차량은 프런트 축 구동 EV이므로, 리제너레이티브 브레이킹 중에 모터에 사용 가능한 프런트 축의 브레이크 힘은 토크로 변환되어 브레이킹 중에 토크 참조값으로 사용됩니다.

 

To analyze the influence of different factors on the displacement of the braking cutoff point, four factors of deceleration rate, vehicle mass, wind speed, and road slope angle were separately tested on the simulation model discussed above using the specifications mentioned in Table I.

For each simulation, the cutoff point was assumed to be the instant where the motor controller dc link current changes its direction from negative to positive during deceleration.

다양한 요인이 브레이킹 절단점 이동에 미치는 영향을 분석하기 위해 브레이킹 절단점 이동에 영향을 미치는 네 가지 요인인 감속률, 차량 질량, 바람 속도 및 도로 경사 각도를 개별적으로 테스트하였습니다, 테스트 된 시뮬레이션 모델의 구체적 방법을 Table 1에 언급되었습니다.

위에서 언급한 사양을 사용하여 이전에 논의한 시뮬레이션 모델을 사용하여 각 시뮬레이션에 대해, 감속 중인 동안 모터 컨트롤러 DC 링크 전류가 음에서 양으로 바뀌는 시점을 절단점으로 가정하였습니다.

 

The results of this study are depicted in Figs. 7–10.

In these plots, a negative motor controller dc link current indicates recharging the battery, and a positive current shows battery discharge.

Based on Fig. 7, increasing the deceleration rate from 2.5 to 3.33 mph/s increases LSCP from 9.7 to 13.5 mph.

Fig. 8 shows that increasing the vehicle mass from 1000 to 1400 kg increases LSCP from 9.7 to 14.2 mph.

While the results of Fig. 9 indicate that a wind speed of 20 m/s in the opposite direction of vehicle movement, compared to no wind conditions, decreases LSCP from 9.7 to 8 mph.

Furthermore, Fig. 10 reveals the effect of a 3° uphill slope in decreasing LSCP from 9.7 to 3.5 mph.

Based on these results, it can be seen that a faster deceleration rate or higher vehicle mass increases the LSCP, while the effect of higher wind speeds and driving uphill results in a decrease in LSCP.

이 연구의 결과는 그림 7-10에 나와 있습니다.

이 그림에서 음수 모터 컨트롤러 DC 링크 전류는 배터리를 충전하는 것을 나타내고, 양수 전류는 배터리 방전을 나타냅니다.

그림 7을 기반으로 감속률을 2.5에서 3.33 mph/s로 증가시키면 LSCP가 9.7에서 13.5 mph로 증가합니다.

그림 8은 차량 질량을 1000에서 1400 kg로 증가시키면 LSCP가 9.7에서 14.2 mph로 증가함을 보여줍니다.

그림 9의 결과는 차량 이동 방향과 반대 방향으로 20 m/s의 바람 속도가 없는 조건과 비교하여 LSCP가 9.7에서 8 mph로 감소함을 보여줍니다.

또한, 그림 10은 3° 오르막 경사가 LSCP를 9.7에서 3.5 mph로 감소시키는 효과를 보여줍니다.

이러한 결과를 통해 더 빠른 감속률이나 더 높은 차량 질량은 LSCP를 증가시키고, 더 높은 바람 속도와 오르막 주행은 LSCP를 감소시키는 효과를 보여줍니다.

 

From these case studies, it is observed that LSCP is dynamically changing due to the constant changes in the operating point of the TM.

Thus, designing a brake controller in which a dynamic LSCP is considered instead of a constant LSCP is a vital step toward maximizing the recaptured energy from the regenerative braking process [31].

이러한 사례 연구에서는 LSCP가 TM의 운영 지점의 지속적인 변화로 인해 동적으로 변화한다는 것을 관찰할 수 있습니다.

따라서 상수 LSCP 대신 동적 LSCP를 고려하는 브레이크 컨트롤러를 설계하는 것이 회생 브레이킹 프로세스로부터 회수된 에너지를 극대화하기 위한 중요한 단계입니다.

 

(4) Proposed EV Brake Controller Considering Dynamic LSCP

The simulation results obtained from Section III indicate that: 1) the point at which regenerative braking is no longer effective is dynamically changing under various circumstances and 2) the direction change in current flow can serve as an accurate indication for dynamically determining LSCP.

Therefore, by designing a brake controller that takes into account the dynamic nature of LSCP, energy loss during regenerative braking can be prevented, and vehicle driving range can be further improved.

In the proposed control strategy, the motor controller dc link current is monitored and fed back to the brake controller to determine the current crossing point and hence identify the dynamic LSCP during regenerative braking.

A flowchart representation of the proposed brake controller is depicted in Fig. 11 that takes advantage of the dc link current as an indicator to accurately determine LSCP.

섹션 III에서 얻은 시뮬레이션 결과는 다음과 같습니다: 1) 리제너레이티브 브레이킹이 더 이상 효과적이지 않은 지점이 다양한 상황에서 동적으로 변화하며, 2) 전류 흐름의 방향 변화가 동적으로 LSCP를 결정하는 정확한 지표로 작용할 수 있습니다. 

따라서 LSCP의 동적 특성을 고려한 브레이크 컨트롤러를 설계함으로써 리제너레이티브 브레이킹 중 발생하는 에너지 손실을 방지하고 차량 주행 거리를 더욱 향상시킬 수 있습니다. 

제안된 제어 전략에서는 모터 컨트롤러 DC 링크 전류를 모니터링하고 브레이크 컨트롤러에 피드백하여 현재가 교차하는 지점을 결정하고 이로써 회생 브레이킹 중 동적 LSCP를 식별합니다.

제안된 브레이크 컨트롤러의 플로우차트 표현은 동적 LSCP를 정확하게 결정하기 위해 DC 링크 전류를 지표로 활용합니다.

 

In the proposed brake controller, the requested TM torque (TTM) is monitored to determine whether the brake pedal is pressed or not.

In this case, a negative TM torque (TTM < 0) represents braking.

If braking is requested, the share of the front axle (Tbf ) and rear axle (Tbr ) brake torques are calculated based on the I curve.

Once the decomposition of brake torque on each axle is known, the brake torque distribution between regenerative-based and frictionbased brakes is implemented based on the dynamic LSCP detection.

In this step, the motor controller dc link current (Idc_Link) is monitored in order to identify the point at which regenerative braking is no longer effective.

If this current is positive while braking (Idc_Link ≥ 0), regenerative braking is disabled, and all the requested brake torque is satisfied through friction braking. Otherwise, the front axle braking share is allocated to regenerative braking, and the rear axle braking share is achieved through the friction braking.

It should be noted that in the proposed controller, the transition between the regenerative braking and friction braking takes place gradually to ensure driver/passenger safety and comfort during braking.

In other words, after LSCP is detected, Tregen is gradually reduced and friction braking is increased with the same rate to guarantee a smooth transition between the two.

제안된 브레이크 컨트롤러에서는 요청된 모터 토크 (TTM)를 모니터링하여 브레이크 페달이 눌려 있는지 여부를 결정합니다. 

이 경우 음의 TM 토크 (TTM < 0)는 브레이킹을 나타냅니다. 

브레이킹이 요청되면, 전륜 (Tbf) 및 후륜 (Tbr) 브레이크 토크의 분배가 I 곡선을 기반으로 계산됩니다. 

각 축에 대한 브레이크 토크 분해가 알려지면, 동적 LSCP 검출을 기반으로 리제너레이티브 및 마찰 브레이크 간의 브레이크 토크 분배가 구현됩니다. 

이 단계에서는 모터 컨트롤러 DC 링크 전류 (Idc_Link)가 모니터링되어 리제너레이티브 브레이킹이 더 이상 효과적이지 않은 지점을 식별합니다. 

이 전류가 브레이킹 중 양수이면 (Idc_Link ≥ 0), 리제너레이티브 브레이킹이 비활성화되고 모든 요청된 브레이크 토크가 마찰 브레이킹으로 충족됩니다. 

그렇지 않으면, 전륜 브레이킹 비율이 리제너레이티브 브레이킹에 할당되고 후륜 브레이킹 비율은 마찰 브레이킹을 통해 달성됩니다. 

제안된 컨트롤러에서는 리제너레이티브 브레이킹과 마찰 브레이킹 사이의 전환이 부드럽게 이루어지도록 조심스럽게 이루어집니다. 

다시 말해, LSCP가 감지된 후 Tregen은 서서히 감소하고 마찰 브레이킹이 동일한 비율로 증가하여 두 가지 사이의 부드러운 전환을 보장합니다.

 

(5) Experimental Test Bench

The experimental HIL setup employed in this paper is shown in Fig. 12 and represented schematically in Fig. 13.

The experimental setup consists of two high-performance PMSMs with their corresponding inverters.

PMSMs have an operating speed of 0–7700 rpm, maximum power of 135 kW, maximum torque of 320 Nm, and operate with input dc voltage range of 270–425 V. The motors can handle maximum input currents up to 500 A.

The motor shafts are connected to each other through a torque sensor that is responsible for accurate measurement of the shaft torque.

In this setup, one of the PMSMs emulates the TM, while the other one acts as the Dyno to mimic road load conditions, vehicle inertia, and braking effects. Real-time communication is achieved using a Controller Area Network (CAN) bus.

LabVIEW software and its real-time capability is utilized for calculating the desired speed and torque commands at each set point.

Emulating vehicle operating conditions and calculation of Dyno torque and TM speed are carried out according to the controller architecture discussed in Section III and the corresponding block diagram that was previously discussed and was shown in Fig. 6.

To achieve synchronous operation of both motors, the TM and Dyno operating points are updated at each step by means of synchronous commands sent from LabVIEW using CAN communication.

To ensure that the trace is always followed, the controller constantly monitors the error between reference signals and the actual measured signals, and if the error is higher than a predetermined margin, the simulation is terminated.

This capability of the controller is for accomplishing precise experimental results under various operating conditions.

Moreover, a 400-V battery cabinet with the capability of bidirectional energy flow is utilized as the main dc source for the test bench that enables energy recapturing during regenerative braking.

이 논문에서 사용된 실험 HIL 설정은 그림 12에 나타나 있으며 도식적으로 그림 13에 표시되어 있습니다. 

실험 설정은 두 개의 고성능 PMSM과 해당 인버터로 구성됩니다. 

PMSM은 0~7700 rpm의 운전 속도, 최대 출력 135 kW, 최대 토크 320 Nm을 갖추고 입력 DC 전압 범위는 270~425 V입니다. 

모터는 최대 500 A까지의 최대 입력 전류를 처리할 수 있습니다. 

모터 축은 서로를 연결하는 토크 센서를 통해 축 토크를 정확하게 측정합니다. 

이 설정에서 하나의 PMSM은 TM을 에뮬레이션하고 다른 하나는 도로 부하 조건, 차량 관성 및 브레이킹 효과를 모방하는 Dyno로 작동합니다. 

실시간 통신은 컨트롤러 에어리어 네트워크 (CAN) 버스를 사용하여 달성됩니다. 

LabVIEW 소프트웨어와 해당 실시간 기능은 각 설정 지점에서 원하는 속도와 토크 명령을 계산하는 데 사용됩니다. 

차량 운전 조건을 모방하고 Dyno 토크 및 TM 속도를 계산하는 작업은 섹션 III에서 설명된 컨트롤러 아키텍처 및 이전에 논의된 대응하는 블록 다이어그램에 따라 수행됩니다(그림 6에 표시됨). 

두 모터의 동기 작동을 달성하기 위해 TM 및 Dyno 운전 지점은 CAN 통신을 통해 LabVIEW에서 보내는 동기화 명령에 따라 각 단계에서 업데이트됩니다. 

추적이 항상 따르도록 하기 위해 컨트롤러는 지속적으로 참조 신호와 실제 측정된 신호 간의 오차를 모니터하고 오차가 미리 결정된 여유보다 크면 시뮬레이션이 종료됩니다. 

이 컨트롤러의 기능은 다양한 운전 조건에서 정밀한 실험 결과를 달성하기 위한 것입니다. 

또한, 회생 브레이킹 중 에너지 회수가 가능한 주요 DC 원으로 400 V 배터리 캐비닛이 사용됩니다.

 

(6) Case Studies and Experimental Results

In order to experimentally verify the effectiveness of the proposed brake control strategy on maximizing the harvested energy, the proposed controller was integrated into the main controller of the experimental test bench.

Two observations were conducted using the experimental test bench and the vehicle data that was previously given in Table I.

제안된 브레이크 제어 전략의 효과를 실험적으로 확인하기 위해, 제안된 컨트롤러를 실험용 테스트 벤치의 주 컨트롤러에 통합했습니다. 

실험 테스트 벤치와 이전에 제공된 차량 데이터를 사용하여 두 가지 관찰을 수행했습니다. 

The first test was carried out using two drive cycles; the Urban Dynamometer Driving Schedule (UDDS); and the New York City Cycle (NYCC) shown in Fig. 14, in which a constant LSCP of 15 mph was considered.

The second test was executed considering the same drive cycle profiles; however, dynamic LSCP was taken into account by implementing the proposed brake controller of Fig. 11.

첫 번째 테스트는 두 가지 운전 주기, 즉 도시 다이나모 주행 일정 (UDDS) 및 뉴욕 시 주기 (NYCC)를 사용하여 진행되었습니다. 

그림 14에 표시된 것처럼, 여기에서는 15 mph의 일정한 LSCP를 고려했습니다. 

두 번째 테스트는 동일한 운전 주기 프로파일을 고려하여 실행되었지만, 동적 LSCP가 그림 11의 제안된 브레이크 컨트롤러를 구현함으로써 고려되었습니다.

 

The amount of harvested energy through the regenerative braking process as well as the required net energy to complete the UDDS (NYCC) drive cycles were calculated and compared for both tests. 

The results are summarized in Table II for the duration of one drive cycle. 

As shown in Table II, the harvested energy during a complete drive cycle is increased by considering dynamic LSCP instead of constant LSCP. 

Thus, the net energy consumption, which is calculated from the difference between the required energy to propel the EV and the recaptured energy by the regenerative braking process, decreases while taking into account the dynamic changing nature of LSCP. 

추적된 에너지 양과 동적 LSCP 대신 상수 LSCP를 고려할 때 완전한 운전 주기 동안 필요한 순 에너지 소비는 표 II에 요약되어 있습니다.

표 II에서 보듯이, 동적 LSCP를 고려함으로써 완전한 운전 주기 동안 수확된 에너지가 증가합니다.

따라서 EV를 추진하는 데 필요한 에너지와 동적으로 변하는 LSCP에 의해 회수된 에너지 사이의 차이로 계산되는 순 에너지 소비는 줄어듭니다.

 

To further analyze the results, the harvested energy during regenerative braking for the complete UDDS drive cycle is depicted in Fig. 15. 

It is evident from Fig. 15 that the recaptured energy increases when considering dynamic LSCP instead of a constant LSCP.

결과를 더 분석하기 위해, 완전한 UDDS 운전 주기 동안 재생 제동 중에 수확된 에너지가 그림 15에 나타나 있습니다.

그림 15에서 동적 LSCP를 고려할 때 재생 제동 중 회수된 에너지가 상수 LSCP를 고려할 때보다 증가하는 것이 명백합니다.

 

Figs. 16 and 17 illustrate the experimental results for both tests for the duration of 20 s of the UDDS drive cycle, during which the vehicle is decelerating from 34.6 mph to a complete stop.

The 20-s sample time was chosen in order to clearly show the difference between the two tests.

Fig. 16 shows the experimental result when a constant LSCP of 15 mph was considered as the low-speed threshold, while Fig. 17 presents the experimental result when a dynamic LSCP was considered as the speed threshold.

그림 16과 17은 UDDS 운전 주기의 20초 동안의 두 실험 결과를 보여줍니다. 이 기간 동안 차량은 34.6mph에서 완전히 멈출 때까지 감속합니다. 

이 20초의 샘플 시간은 두 실험 간의 차이를 명확히 보여주기 위해 선택되었습니다. 

그림 16은 저속 특정 경계로 15mph를 고려했을 때의 실험 결과를 보여주고, 그림 17은 동적 LSCP가 속도 경계로 고려된 실험 결과를 보여줍니다. 

By comparing the results of Figs. 16 and 17, it is noted that since the LSCP in the first experiment was considered fixed, the TM regenerative current falls to 0 as soon as the speed reaches 15 mph.

However, in the second experiment, in which LSCP was dynamically determined by current monitoring, the charging current is extended, and LSCP occurs at 8.9 mph, which is equivalent to zero crossing of the TM controller dc link current.

This extension of current waveform results in an increase in energy harvesting during each braking.

Furthermore, by comparing the results of friction braking shown in Figs. 16 and 17, it is evident that under dynamic LSCP, since regenerative braking is extended, the share of friction brake is less compared to constant LSCP.

그림 16과 17의 결과를 비교하면 첫 번째 실험에서는 LSCP가 고정되어 있기 때문에 속도가 15mph에 도달하자마자 TM 재생 전류가 0으로 떨어진다는 것을 알 수 있습니다. 

그러나 두 번째 실험에서는 LSCP가 전류 모니터링을 통해 동적으로 결정되었으므로 충전 전류가 연장되고, LSCP가 8.9mph에서 발생되며 이는 TM 컨트롤러 dc 링크 전류의 제로 크로싱과 동등합니다. 

이러한 전류 파형의 연장으로 인해 각 제동 시 에너지 수확량이 증가합니다. 

또한 그림 16과 17에 나타난 마찰 제동의 결과를 비교하면 동적 LSCP에서는 재생 제동이 연장되므로 상수 LSCP와 비교하여 마찰 제동의 비율이 적다는 것이 명백합니다.

 

This improvement in energy harvesting is even more significant when taking into consideration that a typical passenger vehicle travels around 12000 mi annually [32].

If one assumes that a vehicle travels daily and solely based on the UDDS (NYCC) pattern, implementing dynamic LSCP instead of constant LSCP can save 3.67% (5.67%) in annual energy.

This translates to 16.75 kWh (55 kWh) of energy per vehicle or the equivalent to almost 17 UDDS (186 NYCC) drive cycles per year.

This means that by implementing the proposed brake control strategy for the vehicle under study in this paper, and considering UDDS (NYCC) as the daily driving profile, an EV owner can extend the annual range of its vehicle by 17 extra UDDS (186 extra NYCC) drive cycles.

Table III summarizes the annual energy extraction and consumption for this vehicle when considering both constant and dynamic LSCP.

It should be noted that this energy saving can be even more significant when considering drive cycles with higher stop-and-go traffic patterns.

Furthermore, the broader impact of the proposed strategy can be realized on a larger scale when considering the increasing number of EVs on the road in the future.

이 에너지 수확의 개선은 년간 주행거리가 약 12000 마일인 일반 승용차를 고려할 때 훨씬 더 중요합니다. 

하루에 한 번 이동하는 차량이 있고 오직 UDDS(NYCC) 패턴에 따라 이동한다고 가정하면, 동적 LSCP를 상수 LSCP 대신 구현함으로써 연간 에너지를 3.67%(5.67%) 절약할 수 있습니다. 

이는 차량당 연간 16.75 kWh(55 kWh)의 에너지를 절약하거나 거의 17개의 UDDS(186개의 NYCC) 주행 주기에 해당합니다. 

따라서 이 논문에서 연구 대상 차량에 제안된 브레이크 제어 전략을 구현하고 UDDS(NYCC)를 일일 주행 프로필로 고려할 때 EV 소유자는 연간 차량 범위를 17개의 추가 UDDS(186개의 추가 NYCC) 주행 주기로 확장할 수 있습니다. 

이 연간 에너지 추출 및 소비를 요약한 테이블 III를 참조하십시오. 

더 중요한 것은 고정된 구간-정지 교통 패턴을 고려할 때 이 에너지 절약이 더욱 중요해질 수 있다는 것입니다. 

또한, 미래에 도로를 달릴 EV의 수가 증가함에 따라 제안된 전략의 더 넓은 영향이 실현될 수 있습니다.

 

(7) Conclusion

In this paper, the dynamically changing nature of the regenerative braking capability of an EV electric motor at low speeds was outlined and different factors influencing this change were studied.

A new control approach based on the motor controller dc link current monitoring was proposed to overcome this limitation and enable dynamic detection of LSCP.

The effectiveness of the proposed control strategy was verified using an experimental motor/Dyno test bench setup.

The results showed that the recaptured energy from the vehicle during braking instances could be significantly improved by means of the proposed control approach.

이 논문에서는 EV 전기 모터의 저속에서 동적으로 변하는 재생 제동 능력에 대한 성격이 강조되었으며, 이 변화를 영향하는 다양한 요소들이 연구되었습니다. 

전동기 컨트롤러 DC 링크 전류 모니터링을 기반으로 한 새로운 제어 접근 방식이 제안되어 이 제한을 극복하고 LSCP의 동적 감지가 가능하게 되었습니다. 

제안된 제어 전략의 효과는 실험적인 모터/다이노 테스트 벤치 설정을 사용하여 검증되었습니다. 

결과는 제안된 제어 접근 방식을 통해 차량의 제동 순간에 회수되는 에너지가 크게 향상될 수 있다는 것을 보여주었습니다.