WO2018104850A1

WO2018104850A1 - Model predictive based control for automobiles

Info

Publication number: WO2018104850A1
Application number: PCT/IB2017/057636
Authority: WO
Inventors: Somnath Sengupta; Chethan Gururaja; Aditya Chandrasekar RAMESH
Original assignee: Kpit Technologies Limited
Priority date: 2016-12-08
Filing date: 2017-12-05
Publication date: 2018-06-14

Abstract

Described herein is a model predictive control based system for a retrofit hybrid electric vehicle (HEV) or electric vehicle (EV). The system comprises a vehicle parameter input receive module to receive parameter inputs, a torque assist value generation module to process the received vehicle parameter inputs using a model predictive control (MPC) technique to generate an optimal torque assist value (torque/torque value in case of EV), where the optimal torque assist value being optimally split into a first torque component that is required for a vehicle engine and a second torque component that is required for vehicle motor in a manner so as to minimize fuel consumption and overall energy consumption while satisfying constraints associated with safety and drivability; and a motor operation module to operate a vehicle motor using a motor torque command generated based on the second torque component.

Description

MODEL PREDICTIVE BASED CONTROL FOR AUTOMOBILES

TECHNICAL FIELD

[0001] The present disclosure relates generally to the field of supervisory controllers in automobiles. In particular, the present disclosure pertains to using a Model Predictive based Control (MPC) methodology in the supervisory controller of automobiles.

BACKGROUND

[0002] Background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.

[0003] Hybrid vehicles offer consumers with an alternative to vehicles employing conventional Internal Combustion (IC) engines, transmissions, and drive trains which often exhibit relatively low fuel efficiency and/or produce undesirable emissions that are released during operation. A typical hybrid vehicle combines a battery or fuel cell powered electric motor with an IC engine. A hybrid vehicle, either factory built or retrofitted, reduces fossil fuel consumption by employing an additional set of prime movers that run on electric power in addition to the IC engine. Although fossil fuel consumption is reduced, it requires coordinated control of the two prime movers to minimize the fossil fuel consumption, which also consequently leads to reduction of the emissions. Further, coordinated control requires knowledge of both prime movers.

[0004] There are several Model Predictive Control (MPC) based supervisory controllers employed in conventional hybrid electric vehicles that coordinate control of all prime movers to minimize fossil fuel consumption. These Model Predictive Controllers employed in factory-built hybrid electric vehicle/electric vehicles have access to all the prime movers and can control them to achieve maximum fossil fuel economy. However, the controller in a retrofit hybrid vehicle/electric vehicle suffers from lack of knowledge of dynamic engine information (lack of control and sensor interfaces) and hence cannot control the IC engine to minimize fossil fuel consumption.

[0005] United States Patent US 8596391 B2 provides a method of converting a vehicle having an Internal Combustion engine, a transmission, an alternator and a battery into a hybrid vehicle, also sometimes referred to as retrofit hybrid vehicle. A Hybrid Electric Vehicle (HEV) is a type of hybrid vehicle and electric vehicle that combines a conventional IC engine propulsion system with an electric propulsion system, the two prime movers constituting hybrid vehicle drivetrain.

[0006] German Patent Application DE 10103188 Al provides a module, having a battery, a motor, a motor controller and a transmission along with the supervisory controller, which is added to an existing conventional IC powered automotive system with typical configuration for converting conventional vehicles into hybrid vehicles.

[0007] The main purpose for such a controller system is to minimize fuel consumption and losses along with reducing stresses on electrical components, using the motor which shares some of the demanded torque by the driver. However, the only variable that can be controlled to achieve the above advantage is the motor torque assist to the engine which eventually leads to torque split between the engine and motor, further aided by the driver in loop phenomena.

[0008] However, the current requirement of the supervisory control for a retrofitted architecture HEV system is to control powertrain in the most optimal manner using MPC by sharing some load with IC engine to achieve maximum benefits mentioned earlier along with respecting system constraints to ensure safety, such as battery voltage/SOC, battery/motor temperature, etc. This is to be done without interacting with the engine but only through the motor torque command and anticipation of appropriate pedal response from the driver, throughout the vehicle operation. Similarly, the current requirement of the supervisory control for a pure EV system is to control power train in the most optimal manner considering optimal operating points of electric components and using MPC by allocating power between tractive torque and various loads to achieve maximum benefits along with respecting system constraints to ensure safety, such as battery voltage/SOC, battery/motor temperature, etc.

[0009] Chinese Patent CN 102019926 B provides predictive energy management control scheme for a vehicle including a hybrid powertrain system. The predictive energy management control scheme provides a method for controlling a vehicle having a hybrid powertrain, the method includes monitoring and forecasting the path of travel related to vehicle navigation and traffic patterns. Powertrain controller will instantaneously expand to the predictive control framework, and the use of traffic based on anticipated vehicle sensing and geographic information and navigation information. Impending road load is predicted, thereby optimizing fuel consumption coefficients in model predictive control framework. [0010] United States Patent US 7360615 B2 provides a predictive energy management system for (non-retrofit) hybrid electric vehicles. The predictive energy management system for a hybrid vehicle that uses certain vehicle information, such as present location, time, 3-D maps and driving history, to determine engine and motor power commands. The system forecasts a driving cycle profile and calculates a driver power demand for a series of N samples based on a predetermined length of time, adaptive learning, etc. The system generates the optimal engine and motor power commands for each N samples based on the minimization of a cost function under constraint equations. The constraint equations may include a battery charge power limit, a battery discharge power limit, whether the battery state of charge is less than a predetermined maximum value, whether the battery state of charge is greater than a predetermined minimum value, motor power output and engine performance. The system defines the cost function as the sum of the total weighted predicted fuel consumed for each sample. The system then selects the motor and engine power commands for the current sample.

[0011] However, the existing supervisory control methodologies for HEV based on MPC determine power distribution coefficient/preferred torque split ratio between engine and motor(s) in a full-fledged HEV system (non-retrofit), where, throughout the operation it is expected that both engine and motor(s) can be sensed using sensors and can be controlled through respective actuators. However, the existing solutions for retrofit HEV/pure EV do not use MPC in supervisory control which is a proven control technology for yielding optimal results in performance.

[0012] The non-patent literature "Development of a retrofit split-axle parallel hybrid electric vehicle with in-wheel motors" in 4th International Conference on Intelligent & Advanced Systems (ICIAS) disclosed in a split-axle parallel hybrid electric vehicle configuration. It uses in-wheel motor to convert existing vehicle into simple retrofitted hybrid configuration. It uses energy management system (EMS) as a main energy controller which may be based on optimal control strategies. This energy controller can only command electric motor. It does not send any signal to internal combustion engine. Motor torque can be varied by using throttle positioning sensor. However, this document does not reveal MPC based algorithm for abnormal conditions, adaptive weightages on cost function for abnormal conditions, display device for indicating gear position, accelerator pedal position to driver. [0013] The non-patent literature "Minimizing Battery Stress during Hybrid Electric Vehicle Control Design" talks about retrofitted hybrid electric vehicle which uses electric motor in the existing conventional Internal Combustion Engine. Dynamic Programming technique has been used to optimize the control strategy over known driving cycle. It reveals cost function (Ref eq. l) which includes additional weighting factor in order to minimize negative effects like battery management, battery state of health, and motor temperature (fault/ abnormal condition).

[0014] The non-patent literature "Split-parallel in-wheel-motor retrofit hybrid electric vehicle" disclosed in Power Engineering and Optimization Conference (PEDCO) Melaka, Malaysia talks about retrofitted hybrid vehicle. It reveals different control strategies for energy management system which also include model predictive control system. It uses driver interface for taking drivers operating mode input and for display of vehicle operating parameter. Electric motor is used to drive IC engine closer to optimal region.

[0015] However, it would be advantageous if an MPC method is able to minimize fossil fuel consumption and hence emissions even under lack of dynamic information of internal combustion engine or ability to directly control the engine. More specifically, it would be advantageous if the provided Model Predictive Control method is not dependent on the access to engine sensors and actuators throughout the entire vehicle operation to achieve the same benefits as a conventional MPC based HEV supervisory controller.

[0016] There is therefore a need for providing an MPC methodology in the supervisory controller of a retrofitted Hybrid Electric Vehicle (HEV) / pure Electric Vehicle (EV) that minimizes fossil fuel consumption/electric energy consumption despite lack of dynamic information of the internal combustion engine or ability to directly control it, resulting in better performance of the vehicle. Also, in case of EV there exist a need to operate and plan/arbitrate the allocation of motor torque, heating, ventilation and air conditioning (HVAC) load, electric loads (lights, wipers, etc.) such that the net electric energy consumption is minimal.

[0017] All publications herein are incorporated by reference to the same extent as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. Where a definition or use of a term in an incorporated reference is inconsistent or contrary to the definition of that term provided herein, the definition of that term provided herein applies and the definition of that term in the reference does not apply. [0018] In some embodiments, the numbers expressing quantities of ingredients, properties such as concentration, reaction conditions, and so forth, used to describe and claim certain embodiments of the invention are to be understood as being modified in some instances by the term "about." Accordingly, in some embodiments, the numerical parameters set forth in the written description and attached claims are approximations that can vary depending upon the desired properties sought to be obtained by a particular embodiment. In some embodiments, the numerical parameters should be construed in light of the number of reported significant digits and by applying ordinary rounding techniques. Notwithstanding that the numerical ranges and parameters setting forth the broad scope of some embodiments of the invention are approximations, the numerical values set forth in the specific examples are reported as precisely as practicable. The numerical values presented in some embodiments of the invention may contain certain errors necessarily resulting from the standard deviation found in their respective testing measurements.

[0019] As used in the description herein and throughout the claims that follow, the meaning of "a," "an," and "the" includes plural reference unless the context clearly dictates otherwise. Also, as used in the description herein, the meaning of "in" includes "in" and "on" unless the context clearly dictates otherwise.

[0020] The recitation of ranges of values herein is merely intended to serve as a shorthand method of referring individually to each separate value falling within the range. Unless otherwise indicated herein, each individual value is incorporated into the specification as if it were individually recited herein. All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g. "such as") provided with respect to certain embodiments herein is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention otherwise claimed. No language in the specification should be construed as indicating any non-claimed element essential to the practice of the invention.

OBJECTS OF THE INVENTION

[0021] It is an object of the present disclosure to provide a supervisory controller for a retrofit HEV that needs no interface with the engine sensors or actuators, but results in pushing engine operation towards optimal region and savings in fuel. [0022] It is another object of the present disclosure to provide an MPC framework which can deal with abnormal/fault conditions as well as reduce jerks/torque pulsations for pure EV as well as retrofit HEV cases.

[0023] It is another object of the present disclosure to provide an MPC framework in case of retrofit HEV which is not dependent on the access to engine sensors and actuators throughout the entire vehicle operation to achieve the same benefits as a conventional MPC based HEV supervisor.

[0024] It is an object of the present disclosure to achieve best possible and reliable performance, while respecting constraints such as safety for a pure EV as well as a retrofitted HEV, using MPC framework.

SUMMARY

[0025] This summary is provided to introduce concepts related to model predictive based control system. The concepts are further described below in the detailed description. This summary is not intended to identity key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0026] In an aspect, the present disclosure relates to a model predictive control based system for a retrofit hybrid electric vehicle (HEV). The system comprises a vehicle parameter input receive module, which when executed by a supervisory controller that forms part of the system, receives, at the supervisory controller, one or more vehicle parameter inputs. The one or more vehicle parameter inputs may be selected from any or a combination of, but not limited to, brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltage, state of charge (SOC) of battery, battery and motor temperature, and battery and motor current. The system further comprises a torque assist value generation module, which when executed by the supervisory controller, generating an optimal torque assist value using a model predictive control (MPC) technique that processes the received one or more vehicle parameter inputs. The optimal torque assist value is optimally split into a first torque component that is required for vehicle engine and a second torque component that is required for vehicle motor in a manner so as to minimize fuel consumption. The system further comprises a motor operation module, which when executed by the supervisory controller, operating a vehicle motor of the retrofit HEV based on the second torque component, using a motor torque command sent by the supervisory controller to a motor controller of the vehicle motor. In an implementation, the motor torque command is optimally split when said retrofit HEV includes a plurality of motors.

[0027] In an implementation, the MPC technique comprises a prediction mechanism that predicts dynamic operational engine variables, based on which the optimal torque assist value is generated. The said predicted dynamic operational engine variables may be selected from any or a combination of vehicle speed, motor power and efficiency, engine power and efficiency, and one or more constraints pertaining to motor and/or battery parameters.

[0028] In an implementation, the MPC technique optimizes generation of the optimal motor torque assist value in abnormal conditions by using adaptive weightages and scaling factors. The MPC technique assigns the adaptive weightages to one or more vehicle parameter inputs. Further, the MPC technique enables computation of the optimal torque assist value based on a cost function factor of predicted fuel consumption.

[0029] In an implementation, the MPC technique is configured to predict present/future torque demands based on one or more inputs selected from any or a combination of road profile in front, road attributes, curves, and gradient.

[0030] In an implementation, the MPC technique is configured to predict any or a combination of present load/comfort requirements and distance mapping.

[0031] In an implementation, the supervisory controller is configured to transmit one or more recommendations to the driver of the retrofit HEV pertaining to any or a combination of optimal gear position and optimal accelerator pedal position so as to minimize the fuel consumption.

[0032] In an implementation, the supervisory controller predicts future dynamics of the retrofit HEV based on the MPC technique. Also, the supervisory controller maximizes regeneration based on the MPC technique.

[0033] In an implementation, the system generates a cost function based on any or a combination of the predicted dynamic operational engine variables, vehicle speed, motor power, motor efficiency, engine power, engine efficiency, and one or more constraints pertaining to any or a combination of motor parameters, battery parameters and/or the predicted dynamic operational engine variables. The cost function is used by said MPC technique to minimize the fuel consumption and overall energy consumption.

[0034] In an implementation, the system is operatively coupled with a post processing block that incorporates the optimal torque assist value so as to minimize jerk/torque pulsation. [0035] In another aspect, the present disclosure further relates to a system for an electric vehicle (EV). The system comprises a vehicle parameter input receive module, which when executed by a supervisory controller that forms part of the system, receiving, at the supervisory controller, one or more vehicle parameter inputs such as vehicle speed, motor current, battery variables, etc. The system further comprises a torque value generation module, which when executed by the supervisory controller, generates an optimal torque value using a model predictive control (MPC) technique that processes the received one or more vehicle parameter inputs. The system further comprises a torque value split module, which when executed by the supervisory controller, optimally splits the optimal torque value into a tractive torque component and a load torque component. The load torque component is used for one or more loads such as HVAC, lights, etc that form part of the EV in a manner so as to minimize energy consumption.

[0036] In an implementation, the MPC technique comprises a prediction mechanism that further predicts future torque demands based on any or a combination of road profile in front, load requirement, and comfort requirement.

[0037] In an implementation for an EV, the system generates a cost function based on any or a combination of motor power, motor efficiency, temperature of motor or battery, regeneration, error in optimal torque or load, and one or more constraints pertaining to any or a combination of motor and battery parameters, acceleration, and torque requirement. The cost function is used by said MPC technique to generate the optimal torque value.

[0038] It may be appreciated that, for the case of pure EV, the MPC framework in the supervisory controller will operate in EV mode to plan/arbitrate allocation of motor torque for tractive purpose (e.g., operate motor close to optimal region), HVAC load, electric loads (lights, wipers, etc.) such that the total electric energy consumption is minimal and regeneration is maximum, while maintaining safe temperature limits for motor, motor controller and battery.

[0039] It is to be noted that, the expression "Torque assist" in terms of HEV indicates an actual assist provided by a motor while primarily an engine drives a vehicle, whereas the expression "torque assist" (wherever used and as far as applicable) in terms of pure EV indicates an actual torque (and not an 'assist') provided by the motor which drives the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0040] The illustrated embodiments of the subject matter will be best understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and processes that are consistent with the subject matter as claimed herein, wherein:

[0041] FIG. 1 illustrates an exemplary block diagram of retrofit hybrid electric vehicle in accordance with an embodiment of the present disclosure;

[0042] FIG. 2 illustrates a graphical map representing a performance comparison for a first exemplary drive cycle in accordance with the present disclosure;

[0043] FIG. 3 illustrates a graphical map representing a performance comparison for a second exemplary drive cycle in accordance with the present disclosure;

[0044] FIG. 4 illustrates an exemplary block diagram showing a calculation of an engine revolution per minute (RPM) based on the vehicle speed in accordance with an embodiment of the present disclosure;

[0045] FIG. 5 illustrates an exemplary block diagram showing a calculation of the engine torque based on vehicle acceleration in accordance with an embodiment of the present disclosure;

[0046] FIG. 6 illustrates behavior of the MPC under the normal operating condition or abnormal operating condition in accordance with an embodiment of the present disclosure;

[0047] FIG. 7 illustrates an exemplary block diagram showing generation of optimal torque and suggested values of optimal gear and throttle command to be displayed to the user by the MPC in accordance with an embodiment of the present disclosure;

[0048] FIG. 8 illustrates a block diagram of the MPC based supervisory controller for only EV case in accordance with an embodiment of the present disclosure;

[0049] FIG. 9 illustrates a flow diagram representing an exemplary implementation of optimization routine in a MPC controller in accordance with an embodiment of the present disclosure;

[0050] FIG. 10 illustrates a functional structure of the MPC in accordance with an embodiment of the present disclosure; and

[0051] FIG. 11 illustrates a method to port complex MPC code structure into simple hardware in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0052] The following is a detailed description of embodiments of the disclosure depicted in the accompanying drawings. The embodiments are in such detail as to clearly communicate the disclosure. However, the amount of detail offered is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure as defined by the appended claims.

[0053] As used in the description herein and throughout the claims that follow, the meaning of "a," "an," and "the" includes plural reference unless the context clearly dictates otherwise. Also, as used in the description herein, the meaning of "in" includes "in" and "on" unless the context clearly dictates otherwise.

[0054] Each of the appended claims defines a separate invention, which for infringement purposes is recognized as including equivalents to the various elements or limitations specified in the claims. Depending on the context, all references below to the "invention" may in some cases refer to certain specific embodiments only. In other cases, it will be recognized that references to the "invention" will refer to subject matter recited in one or more, but not necessarily all, of the claims.

[0055] Embodiments of the present invention may be provided as a computer program product, which may include a machine-readable storage medium tangibly embodying thereon instructions, which may be used to program a computer (or other electronic devices) to perform a process. The machine-readable medium may include, but is not limited to, fixed (hard) drives, magnetic tape, floppy diskettes, optical disks, compact disc read-only memories (CD-ROMs), and magneto-optical disks, semiconductor memories, such as ROMs, PROMs, random access memories (RAMs), programmable read-only memories (PROMs), erasable PROMs (EPROMs), electrically erasable PROMs (EEPROMs), flash memory, magnetic or optical cards, or other type of media/machine-readable medium suitable for storing electronic instructions (e.g., computer programming code, such as software or firmware).

[0056] Various methods described herein may be practiced by combining one or more machine-readable storage media containing the code according to the present invention with appropriate standard computer hardware to execute the code contained therein. An apparatus for practicing various embodiments of the present invention may involve one or more computers (or one or more processors within a single computer) and storage systems containing or having network access to computer program(s) coded in accordance with various methods described herein, and the method steps of the invention could be accomplished by modules, routines, subroutines, or subparts of a computer program product.

[0057] Groupings of alternative elements or embodiments of the invention disclosed herein are not to be construed as limitations. Each group member can be referred to and claimed individually or in any combination with other members of the group or other elements found herein. One or more members of a group can be included in, or deleted from, a group for reasons of convenience and/or patentability. When any such inclusion or deletion occurs, the specification is herein deemed to contain the group as modified thus fulfilling the written description of all groups used in the appended claims.

[0058] In the description herein, the present disclosure pertains generally to technical field of optimization based supervisory controllers in automobiles with electrified powertrain. Specifically, the present disclosure pertains to using Model Predictive based Control (MPC) technique in a supervisory controller of a retrofit Hybrid Electric Vehicle (HEV) and/or Electric Vehicle (EV).

[0059] Accordingly, various embodiments and/or implementations of the present subject matter disclosed herein relates to model predictive control based system for a retrofit hybrid electric vehicle (HEV) or Electric Vehicle (EV). In an aspect, the present disclosure relates to a model predictive control based system for a retrofit hybrid electric vehicle (HEV). The system comprises a vehicle parameter input receive module, which when executed by a supervisory controller that forms part of the system, receiving, at the supervisory controller, one or more vehicle parameter inputs. In an example, the one or more vehicle parameter inputs may be selected from any or a combination of brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltage, state of charge (SOC) of battery, battery and motor temperature, and battery and motor current.

[0060] The system further comprises a torque assist value generation module, which when executed by the supervisory controller, generating an optimal torque assist value using a model predictive control (MPC) technique that processes the received one or more vehicle parameter inputs. The optimal torque assist value is optimally split into a first torque component that is required for vehicle engine and a second torque component that is required for vehicle motor in a manner so as to minimize fuel consumption. Here, the MPC technique is not data driven, but physics/vehicle dynamic based logic is used in supervisory controller for a retrofit hybrid electric vehicle. Also, the MPC technique does not generate the "preferred/optimal power split ratio", rather it only generates an optimal motor torque assist value (supervisory controller does not know/decide the ratio of split- it is provided indirectly based on driver torque requirement/pedal feedback loop). In an example, the MPC technique optimizes the generation of the optimal motor torque assist value in abnormal conditions by using adaptive weightages and scaling factors. Accordingly, the MPC technique considers abnormal conditions such as fault conditions and not driving conditions to generate MPC based adaptive motor torques and weights of cost function factor during such abnormal conditions. Further, in an example, the MPC technique assigns the adaptive weightages to one or more vehicle parameter inputs, and enables computation of the optimal torque assist value based on a cost function factor of predicted fuel consumption.

[0061] The system further comprises a motor operation module, which when executed by the supervisory controller, operating a vehicle motor of the retrofit HEV based on the second torque component, using a motor torque command sent by the supervisory controller to a motor controller of the vehicle motor. In an implementation, the motor torque command optimally splits when said retrofit HEV includes a plurality of motors. Accordingly, the motor operation module iteratively (driver in loop) provides appropriate motor torque command to the motor controller to drive motor operating point of nearest optimal region.

[0062] Thus, the disclosed model predictive control based system for the retrofit HEV minimizes fossil fuel consumption despite lack of dynamic information of internal combustion (IC) engine or ability to directly control it.

[0063] Further, with the implementation of the above disclosed aspect of the system, the MPC technique in the supervisory controller of the retrofit HEV results in sharing load, between an IC engine and a battery, in an optimal manner while respecting system constraints to achieve performance enhancement. The present implementation is achieved without interacting with the IC engine but only through the motor torque command, throughout the vehicle operation.

[0064] Yet further, the MPC based supervisory controller of the present disclosure for the retrofit HEV, as compared to the existing control algorithms conventionally implemented by the control methodologies for the retrofit HEV based on MPC, achieves best possible and reliable performance while respecting constraints (ensuring safety) in a formal manner (MPC framework) for a retrofit HEV. The MPC based supervisory controller of the present disclosure needs no interface with the IC engine sensors or actuators, but results in pushing IC engine operation and vehicle motor operation towards optimal region and savings in fuel. So, at each instant of execution step of the MPC based supervisory controller of the present disclosure, the optimal motor torque assist value generated by the supervisory controller corresponds to the best possible reduction in fuel consumption and emissions and overall energy consumption at the same time enable the disclosed system to be limited within the constraints.

[0065] In an implementation, the MPC technique comprises a prediction mechanism that predicts dynamic operational engine variables, based on which the optimal torque assist value is generated. The said predicted dynamic operational engine variables is selected from any or a combination of vehicle speed, motor power and efficiency, engine torque, power and efficiency, and one or more constraints pertaining to motor and/or battery parameters.

[0066] In an implementation, the MPC technique is configured to predict present/future torque demands based on one or more inputs selected from any or a combination of road profile in front, road attributes, curves, and gradient.

[0067] In an implementation, the MPC technique is configured to predict any or a combination of present load/comfort requirements and distance mapping.

[0068] In an implementation, the supervisory controller is configured to transmit one or more recommendations to driver of the retrofit HEV pertaining to any or a combination of optimal gear position and optimal accelerator pedal position so as to minimize the fuel consumption.

[0069] In an implementation, the supervisory controller predicts future dynamics of the retrofit HEV and/or pure EV based on the MPC technique. Also, the supervisory controller maximizes regeneration based on the MPC technique.

[0070] In an implementation, the system generates a cost function based on any or a combination of the predicted dynamic operational engine variables, vehicle speed, motor torque, power, motor efficiency, engine power, engine efficiency, and one or more constraints pertaining to any or a combination of motor parameters, battery parameters and/or predicted dynamic operational engine variables. The cost function issued by said MPC technique to minimize the fuel consumption and overall energy consumption. [0071] In an implementation, the system is operatively coupled with a post processing block that incorporates the optimal torque assist value so as to minimize jerk/torque pulsation.

[0072] In an implementation, the system implements the MPC technique to carry out supervisory controller functions without access to Controlled Area Network (CAN) bus messages of said retrofit HEV.

[0073] In another aspect of the present disclosure, the present disclosure further relates to system for an electric vehicle (EV). The system comprises a vehicle parameter input receive module, which when executed by a supervisory controller that forms part of the system, receiving, at the supervisory controller, one or more vehicle parameter inputs. The system further comprises a torque assist value generation module, which when executed by the supervisory controller, generates an optimal torque t value using a model predictive control (MPC) technique that processes the received one or more vehicle parameter inputs. The system further comprises a torque assist value split module, which when executed by the supervisory controller, optimally splitting the optimal torque/torque value into a tractive torque component and a load torque component. The load torque component is used for one or more loads that form part of the EV in a manner so as to minimize energy consumption.

[0074] In an implementation, the MPC technique comprises a prediction mechanism that further predicts future torque demands based on any or a combination of road profile in front, load requirement, and comfort requirement.

[0075] In an implementation for a pure EV, the system generates a cost function based on any or a combination of motor torque, motor power, motor efficiency, temperature of motor or battery, regeneration, error in optimal torque or load, and one or more constraints pertaining to any or a combination of motor and battery parameters, acceleration, and torque requirement. The cost function is used by said MPC technique to generate the optimal torque/torque value.

[0076] In an aspect, the present disclosure provides MPC technique for use in a retrofit HEV architecture where only vehicle/electric motor can be controlled.

[0077] In the description herein after, various embodiments, implementations, and aspects are further described herein with reference to the accompanying figures. It should be noted that the description and figures relate to exemplary embodiments, implementations, or aspects, and should not be construed as a limitation to the subject matter of the present disclosure. It is also to be understood that various arrangements may be devised that, although not explicitly described or shown herein, embody the principles of the subject matter of the present disclosure. Moreover, all statements herein reciting principles, aspects, implementations, and embodiments of the present disclosure, as well as specific examples, are intended to encompass equivalents thereof. Yet further, for the sake of brevity, operation or working principles pertaining to the technical material that is known in the technical field of the present disclosure have not been described in detail so as not to unnecessarily obscure the present disclosure.

[0078] It is to be noted that, the expression "Torque assist" in terms of HEV indicates an actual assist provided by a motor while primarily an engine drives a vehicle, whereas the expression "torque assist" (wherever used and as far as applicable) in terms of pure EV indicates an actual torque (and not an 'assist') provided by the motor which drives the vehicle.

[0079] FIG. 1 illustrates a block diagram of a model predictive control based system 100 for a retrofit hybrid electric vehicle (HEV) in accordance with an aspect of the present disclosure. As shown in FIG. 1, the system 100 comprises a supervisory controller 102, a driving recommendation unit 104, a vehicle motor(s) 106, a battery pack 108, a brake system 110, a conventional internal combustion (IC) engine 1 12, a clutch and gear box assembly 114, a torque coupling device 116 retrofitted to the IC engine 112, a differential mechanism 118, and wheels 120 of the retrofit HEV.

[0080] In an implementation, the supervisory controller 102 provides a model predictive control based approach that can be implemented with less computational resources and/or with greater speed than other existing approaches. In said implementation, the supervisory controller 102 includes one or more processor(s) 124 and a processing engine(s) 126.

[0081] The one or more processor(s) 124 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that manipulate data based on operational instructions.

[0082] The processing engine(s) 126 may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 126. In example described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processing engine(s) 126 may be processor executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing engine(s) 126 may include a processing resource (for example, one or more processors), to execute such instructions. In the present example, the machine-readable storage medium may store instructions that, when executed by the processing resource, implement the processing engine(s) 126. In said example, the supervisory controller 102 may include the machine-readable storage medium storing the instructions and the processing resource to execute the instructions, or the machine-readable storage medium may be separate but accessible to supervisory controller 102 and the one or more processor(s) 124. In other example, the processing engine(s) 126 may be implemented by electronic circuitry.

[0083] In an implementation, the processing engine(s) 126 comprises a vehicle parameter input receive module 128, a torque assist value generation module 130, and a motor operation module 132. Although the modules 128, 130, and 132 are shown as a part of the supervisory controller 102; however, these modules can be disposed outside the supervisory controller 102 and operatively connected/coupled to the supervisory controller 102, without deviating from the scope of the present disclosure.

[0084] In operation, the vehicle parameter input receive module 128 is configured to receive one or more vehicle parameter inputs 134. In an example, the one or more vehicle parameter inputs 134 can be selected from any or a combination of brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltages, state of charge (SOC) of battery, battery and motor temperature, and battery and motor currents. In said example, the one or more vehicle parameter inputs 134 are received form one or more sensors disposed in the retrofit HEV.

[0085] Once the one or more vehicle parameter inputs 134 are received, the torque assist value generation module 130 processes the received one or more vehicle parameter inputs 134 using a model predictive control (MPC) technique to generate an optimal torque assist value. In an example, the MPC technique comprises a prediction mechanism that predicts (vital) dynamic operational engine variables, based on which the optimal torque assist value is generated. Using these dynamic engine variables, extra cost function factor of predicted fuel consumption is added in the MPC technique for optimization. The predicted dynamic operational engine variables can be selected from any or a combination of vehicle speed, motor power and efficiency, engine power and efficiency, and one or more constraints pertaining to motor and/or battery parameters. [0086] In an alternative implementation, the MPC technique optimizes the generation of the optimal motor torque assist value in abnormal conditions by using adaptive weightages and scaling factors. In said implementation, the MPC technique assigns the adaptive weightages to one or more vehicle parameter inputs, and the MPC enables computation of the optimal torque assist value based on a cost function factor of predicted fuel consumption. In an aspect, the system 100 generates a cost function based on any or a combination of the predicted dynamic operational engine variables, vehicle speed, motor power, motor efficiency, engine power, engine efficiency, and one or more constraints pertaining to any or a combination of motor and battery parameters the predicted dynamic operational engine variables, where said cost function is used by said MPC technique to minimize the fuel consumption and overall energy consumption.

[0087] Continuing with the present disclosure, when the optimal torque assist value is generated, the optimal torque assist value is optimally split into a first torque component that is required for vehicle engine and a second torque component that is required for vehicle motor in a manner so as to minimize fuel consumption. In an example, the motor torque command is further optimally split if said HEV comprises a plurality of vehicle motors.

[0088] Accordingly, based on the second torque component, the motor operation module 132 operates the vehicle motor 106 using a motor torque command sent by the supervisory controller 102 to a motor controller of the vehicle motor 106.

[0089] In one implementation, based on the first torque component of the optimal torque assist value, the supervisory controller 102 is configured to transmit one or more recommendations to driving recommendation module 104, which at its user interface displays recommendation for a driver of the retrofit HEV. The one or more recommendations to driver of the retrofit HEV include any or a combination of optimal gear position (increase/decrease from current position) and optimal accelerator pedal position (increase/decrease from current position) so as to minimize the fuel consumption. Upon receipt and based on the one or more recommendations from the supervisory controller 102, the driver can control the IC engine 112 by operating a brake pedal of the brake system 110, accelerator pedal, and clutch and gear level position by means of clutch and gear box assembly 114. The one or more recommendations can therefore help the driver to operate the retrofit HEV in the most efficient manner for generating a required torque with the IC engine 112. [0090] The torque generated by the IC engine 112 and the torque generated by the vehicle motor 106 are fed to the torque coupling device 116 for being forwarded to a differential mechanism 118. The differential mechanism 118 controls the wheels 120 of the retrofit HEV. The operation of the wheels 120 are connected to the batteries or battery pack 108 of the retrofit HEV. The consumption of the charge from the battery pack 108 can be dependent on the usage of retrofit HEV. Also, the operation of the wheels 120 controls the speed of the retrofit HEV based on MPC technique of the supervisory controller 102. Accordingly, one skilled in the art can appreciate that the disclosed MPC technique based supervisory controller 102 does not generate the "preferred/optimal power/torque split ratio", rather it only generates an optimal motor torque assist value (based on the MPC technique) throughout the entire retrofit HEV operation to result in the same benefits such as fuel savings, maintaining battery limits, and so forth.

[0091] Further, in the system 100 represented in FIG. 1, the disclosed MPC technology in the supervisory controller 102 neither requires nor give any actuation command to the IC engine 112. Rather, the optimal motor torque command corresponding to the first torque component of the optimal torque assist value is given in such a way that it manifests as a torque assist to the retrofit HEV's system so that the driver of the retrofit HEV automatically adjusts the net torque/pedal command, thereby creating a proportion of torque share for each of the IC engine 112 and the vehicle motor 106.

[0092] In one aspect, the supervisory controller 102 for a retrofit HEV architecture is considerably different from that of existing MPC based HEV supervisory controller having different sets of inputs, outputs, constraints and the logic for implementation. Further, the variables of the retrofit HEV system 100 formulated in terms of MPC framework and the corresponding logic developed is executed in a supervisory controller 102 to achieve best possible and reliable performance. In one implementation of MPC framework based control, the variables comprises, but not limited to:

a) Dynamic Model of retrofitted HEV (states) and associated variables

b) Inputs

c) Outputs

d) Constraints

e) Cost Function (to be optimized to yield optimal inputs to achieve the goals) f) Set points / reference trajectory [0093] In this instance of implementation methodology, the following steps have led to a successful implementation of MPC in retrofit HEV:

Incorporating Retrofit HEV Structure into MPC

[0094] The dynamic components and attributes of a retrofit HEV structure are moulded into a typical MPC structure, as follows:

a. Input/manipulated variables: Motor torque

b. States: Vehicle speed, Battery SOC

c. Outputs: Vehicle speed, Battery Current, Battery Voltage

d. Setpoints: Vehicle speed profile

e. Constraints: Battery SOC, Battery Current to be within limits

f. Cost function: To maximize motor usage and minimize overall energy consumption

[0095] To realize a functioning of MPC based supervisory controller for a retrofit HEV, various stages of development are involved. First stage includes developing a physics based dynamic model which should represent and match the dynamics of an actual retrofit HEV as close as possible. Since linear MPC is considered, an algorithm needs a linear state space representation of a plant model. Hence, a second stage would be to linearize and obtain A, B, C & D matrices of the plant model for all desired operating points. These matrices are then fed to the MPC routine which, based on the optimization of formulated cost function, derives an input to be commanded to the plant.

[0096] The generalized dynamic equation for the nonlinear retrofit HEV state space system is written as in equation 1-2:

¾ = ω

y = g(x, u) (2)

where, x, u and y are the states, inputs, and outputs, respectively. Using the nonlinear physics based state space model of the retrofit system, the A, B, C & D matrices are analytically obtained by the Jacobians realized through partial derivatives at the concerned operating points, as in equation 3.

3x dxj 3u du_m

du_m.

[0097] These matrices are functions of states and inputs which are used by MPC routine during runtime. The linearized model realized through A, B, C, and D matrices are to be validated against the non -linear plant model.

[0098] Input to the plant model is a motor torque. Further, throttle position value, brake command and gear command from the driver are considered as exogenous inputs which are not controlled by the supervisory controller but affect the plant dynamics. Further, states of the plant model are vehicle speed and state of charge (SOC).

[0099] A linearized model structure representing the plant dynamics is described by equations 4 and 5 : x(k + l) = Ax(k) + B₁u₁(k) + B₂u₂ {k) (4) y(k) = Cx(k) + Du(k) (5) where, u_x is the manipulated input and u₂ is the set of exogenous inputs. Since there is no feed-forward disturbance, matrix D is considered zero. Using equation 4 and 5, future values of outputs are predicted iteratively as, y(k + 1) = C[Ax(k) + Bu(k)] (6)

[00100] In order to assign weights for errors and tracking the outputs, the state equations must be augmented with the same. The error for the next instant is given by equation 7: e(fc + 1) = [r(fc + 1) - y(k + 1)] (7)

[00101] Considering a causal system, the reference r(k)is taken to be same as the previous value (r(k + 1)) by which equation 8 is realized: e(fc + 1) = e(fc) + y(k) - CAx(k) + CBu(k) (8) [00102] Let n be the prediction horizon. For the n instances, the future state vector and future output vector equations are derived through iterative predictions and are expressed as in equation 9: x_k+n = T z(O) + St/χ + VU₂ (9)

where,

x(0)is the initial value of the state vector.

where A¹ (for /^' =1, . . .«) refers to the power of A corresponding to each succeeding iteration. The U-L is a set of future manipulated inputs over the horizon and U₂ts an array of exogenous inputs at the current instance repeated over the prediction horizon. The generalized cost function for the MPC is given by equation 10.

/O(0), it) = x^T(k)Qx(k) + u^T(k)Ru(k) + x^T(N)Px(N) (10) where, Q is weight matrix for the current and future states and R is weight matrix for the current and future inputs. The cost function of MPC is realized as in equation 1 1.

/O(0), u) = x^T(0)Qx(0) + [Tx(0) + SU_± + VU₂]^TQ [Tx(0) + SU_± + VU₂] + U₁ ^TRU₁

(11)

[00103] By expanding equation 1 1, the MPC cost function can be written in quadratic form so that standard optimization techniques are applicable for its minimization. The cost function is optimized in runtime to obtain the optimal control input over the prediction horizon. The generic form of cost function is presented in equation 12:

/O(0), u) = ^ U₁ ^THU₁ + x' 0 FU₁ + U₂'GU_± + Constant (12) where,

H = 2[R + S'QS]

F = 2[T'QS]

G = 2[V'QS]

[00104] The 'constant' term on the right-hand side of equation (12) is the sum of terms which are independent of the control input.

[00105] The cost function formulated by vehicle speed and SOC as states, error of outputs with respect to the references and outputs as its terms may be realized by A, B, C, D matrices and weights assigned to states and outputs. The objective of the cost function is to maximize motor operation and reduce vehicle speed error (against drive cycle profile). The weights may be chosen such that the motor torque command minimizes fuel consumption and maximizes battery life.

[00106] Yet further, voltage measurement and motor speed (indirectly calculated from the vehicle speed measurement) may be incorporated in equation relating battery current with motor torque to obtain limiting torque value corresponding to the maximum specified current. This calculated limiting torque value may be used as a constraint in optimization. The MPC disclosed above may be required to maintain SOC within the specified limits. This is achieved by constraining the states in the cost function.

Porting of MPC code into existing MPC controller/new controller

[00107] In an embodiment, a method to port a complex MPC code structure into a simple hardware for real time control operations in retrofit HEV vehicles is as described below. The proposed method for porting MPC code is applicable for existing systems as well as for new systems with certain modifications, as will be noted by people skilled in the art.

[00108] In an implementation, the method can include the step of reducing the prediction horizon (based on operating tractive dynamic responses) in the MPC implementation to reduce the memory consumed in RAM, optimizing the memory occupied by the code of MPC, removing continuous/iterative/repetitive calculations of states in the system, and using optimization routine(s) during the implementation of MPC.

[00109] In an exemplary implementation, reducing the prediction horizon (based on operating tractive dynamic responses) in the MPC implementation, reduces the memory consumed in RAM and thereby reduces dimension of resultant prediction matrices. Prediction horizon is the amount of time, states are being predicted. If prediction horizon is decreased more than expected then it is not appreciated. So prediction horizon maybe reduced in a way it doesn't affect the output. One exemplary way to perform the calculation is, for N is the prediction horizon, it will be of multiple matrices calculation of dimension NxN, so smaller the N lesser the size consumed.

[00110] In an exemplary implementation, the memory occupied by the code of MPC is optimized, for example by splitting of functions with suitable memory sizes, fitting into memory map of processor, converting to macros for iterative calls in ROM and further optimizing the RAM consumption (e.g. changed data types, minimizing dynamic variables in main function and alternatively sending across called macros). Other ways of optimization that may be used are using generic methods like writing function, macro, assigning data types and splitting and memory mapping and dynamic memory allocation where also used to optimize memory and RAM consumption.

[00111] In an exemplary implementation, the continuous/iterative/repetitive calculations of states in the system are removed by using pre-formulated states expression (function of instantaneous states and inputs) and adding discrete calculation for state resulting in reduced computation and complexity of calculation which also reduces memory usage. Differential state equations are the pre-formulated differential states expressions which are used to predict the future possible outcomes.

[00112] In an exemplary implementation, the optimization routine(s) are used during the implementation of MPC. Active sets are also hard coded and created as a separate macro and is also further optimized instead of using the of-the shelf inbuilt/standard optimization function to ensure that computation complexity and memory usage is reduced. Experiments and Results

[00113] In an implementation, when the disclosed MPC based supervisory controller with a validated retrofit HEV plant model is executed in a simulation environment to achieve the objective of enhanced performance while respecting the constraints, a comparison is drawn with real data from the conventional controller, while taking the same inputs as that of the conventional controller.

[00114] For instance, for the drawing comparison, conventional retrofit F£EV system - which uses a rule based supervisory controller logic to provide motor torque command - can be a successfully functional vehicle. The data acquired from actual vehicle during operation under KDC and DDC is analysed and the results are retrieved. From the retrieved results, analysis of the behaviour of motor torque, SOC, vehicle speed, battery voltage and current for both the controllers under same input conditions, may enable objective evaluation of the supervisory controller.

[00115] Further, to understand the performance of the MPC based supervisory controller for retrofit F£EV, closed loop simulation of the same vehicle system with MPC based supervisory controller can be executed and the behaviour of the same set of variables may be analysed for the considered drive cycles. The results obtained are compared against those from actual drive cycle data captured from rule based supervisory controller.

[00116] The performance comparison of the two controllers during KDC and DDC are captured by plots which are represented in FIGS. 2 and 3, respectively. Also, their electric energy consumption and regeneration may be compared under the same drive cycles to infer on the extent of motor operation leading to energy balance in the battery.

[00117] From FIGs, 2 and 3, it is evident to a person skilled in the art that the motor torque contribution (during motoring and regeneration) of the plant using MPC based logic is overall greater than that of the conventional case, while following the desired drive cycle speed profile and maintaining the prescribed limits for battery current, SOC and voltage.

[00118] Higher motoring torque (positive in the plot) corresponds to lower proportion of engine usage leading to possible reduction of fuel and emissions and greater regeneration torque (negative in the plot) enables higher energy feedback to battery.

[00119] FIG. 4 illustrates a block diagram 400 showing calculation of an engine revolution per minute (RPM) 406 based on the vehicle speed 402 in case of a retrofit hybrid electric vehicle (HEV). As shown in FIG. 4, by the knowledge of the parameters of transmission components such as gear box, torque couplers, and the like, the corresponding engine speed is mapped from the vehicle speed. In one example, as shown in FIG. 4, the engine RPM 406 can be calculated based on the vehicle speed 402 by using a physics based equation 404 that is pre-configured / pre-stored in a retrofit HEV system.

[00120] FIG. 5 illustrates a block diagram 500 showing a calculation of the engine torque based on vehicle acceleration. As shown in FIG. 5, by predicting the torque and speed of engine, the MPC technique removes the need for sensing them through the engine. This yields the current operating point of the engine (and corresponding fuel consumption, if having a BSFC map of engine). In one example, for the retrofit HEV, using the derived expression for engine torque and speed along with BSFC map, the corresponding derived expression for fuel consumption is added in the cost function of MPC to realize its optimization. This is further used by the disclosed MPC based supervisory controller to iteratively provide appropriate torque command to motor controller which will eventually lead to shifting the operating point of engine towards the nearest optimal one or towards desired of operating point (e.g., lower, higher).

[00121] In one example, as shown in FIG. 5, for the retrofit HEV, the torque at the wheel 506 can be based on a physics based equation 504 calculated using the vehicle acceleration 502. The torque at engine 512 are based on gear box dynamics 510 which can be calculated as a subtract effect of motor torque (in common transmission) 508 which can be based on the torque at the wheel 506.

[00122] FIG. 6 illustrates behaviour 600 of the MPC under the normal operating condition or abnormal operating condition for the retrofit HEV as well as pure electric vehicle (EV). As shown in FIG. 6, based on the type of the abnormal operating condition (e.g., overheating of battery/motor, below threshold SOC, high torque pulsation, etc.), to enable continuation of the motor operation, the weights of the cost function will be changed along with appropriate adaptive scaling of motor/torque command (being conveyed to the motor controller).

[00123] In one example, as shown in FIG. 6, the MPC technique can be adapted to operate in two fault modes 602. First fault mode 602 is a normal condition 604 in which the MPC technique generate functions having fixed weight and no scaling factor for motor torque command. Second fault mode 602 is an abnormal condition 606, wherein based on the type and intensity of the abnormality the MPC can generate adaptive weights on cost functions or modified cost functions or adaptive scaling factor for motor torque command. [00124] FIG. 7 illustrates a block diagram 700 for retrofit HEV, showing generation of optimal torque and suggested values of optimal gear and throttle command to be displayed to the user by MPC technique.

[00125] In one example, as shown in FIG. 7, the MPC technique 704 implemented under a supervisory controller is configured to receive inputs 702 from plurality of sources. In said examples, the inputs 702 may comprise of any or a combination of brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltages, state of charge (SOC) of battery, battery and motor temperatures, and battery and motor currents.

[00126] Further, in FIG. 7, TPS is representing throttle position sensor (value), SOC is representing state of charge, T_qm is representing motor torque, T_e is representing engine torque, PP is representing post processing, and P indicates physical quantity.

[00127] Yet further, in an example not shown in FIG. 7, the MPC technique 704 includes an algorithm which after every successful optimization generates the following optimized values:

(a) Optimal total torque command only to the actuator(s) of motor(s),

(b) If multiple motors, the total optimal motor torque command is further optimally split (considering optimal operating regions of each motor) among each motor, and

(c) Suggested values of optimal gear and throttle command for display (to possibly shift engine operation towards optimal operating point).

[00128] In an implementation, the MPC technique based logic, valid for the retrofit HEV as well as pure EV, generates commanded T_qm to minimize jerk/torque pulsation to enhance drivability, sensing of TPS rate (as input) to understand the requirement of driver in order to command the appropriate torque command which helps to smoothly achieve the net torque along with optimizing other factors (in cost function). In said implementation, the "Post Processing" block 706, valid for the retrofit HEV as well as pure EV, incorporates the optimal torque assist value (or torque/torque value in cases of EV) generated by the MPC technique 704 so as to minimize jerk/torque pulsation by limiting the optimal inputs generated by the MPC technique based supervisory controller, based on the operating condition. For example, the post processing block 706 can limit the rate of commanded motor torque to reduce undesired pulsations in motor current or to reduce jerk.

[00129] In another implementation, since there is no intervention by supervisory controller, IC engine 716 can respond exactly as per the TPS command given by driver 714 in a mostly proportional manner. However, the supervisory controller may or may not give torque command to vehicle motor 710 as per the TPS. For example, with lowering of the TPS, the delivered engine torque decreases but the supervisory controller, being intelligent, gives increasing command to motor controller 708 as per need to optimize cost function to achieve a desired performance/target.

[00130] In another implementation, in case of cruise control, an optimal torque by the vehicle motor 710 drives the IC engine 716 towards the nearest optimal operating point, at the same time maintaining speed (as per the set-point given).

[00131] In another implementation, the torque coupler 712 combines the torque generated by the motor Tqm (P) and the torque generated by the engine Te (P) to adjust the overall torque to operate and/or control the vehicle 718.

[00132] Further, in one implementation, to extend capability of the present subject matter and to address problems of EV as well, additional updated logic can be added for pure EV mode. Since in EV mode, there is no requirement of minimizing engine fuel consumption or reducing emissions, the problem remains to operate and plan/arbitrate allocation of motor torque, heating ventilation and air conditioning (HVAC) load, electric loads (lights, wipers, etc.) such that the net electric energy consumption is minimal and regeneration is maximum, while maintaining safe temperature limits for motor, motor controller, and battery. Additionally, for a given optimal total motor torque command (which leads to minimization of overall electric energy consumption), if multiple motors are present, this motor torque command is further optimally split (considering optimal operating regions of each motor) among each motor.

[00133] To minimize the net electric energy consumption, MPC technique based supervisory controller according to the present disclosure can operate the tractive motor close to optimal region of motor (by choosing the most optimal path to reach the desired torque) as well as give preferences on comfort (e.g., HVAC) leading to dynamically varying weighted minimization of cost function that, for example, can be total energy consumption, to make a decision that can result in an optimal split between tractive torque and other loads.

[00134] To assist the net electric energy consumption to be minimal, the MPC technique based supervisory controller according to the present disclosure can predict future dynamics of vehicle such as speed, cabin temperature, component temperature etc., and based on the predicted dynamics suggestions for compromising on performance can be made to driver which can optimally strike a balance between energy savings and comfort, with comfort having higher priority. For example, reduced torque allocation to accommodate HVAC activity can be suggested which may lead to reaching a higher intended velocity at a delayed time.

[00135] To ensure that net electric energy consumption in EV is minimal, the MPC based supervisory controller according to the present disclosure can be used to anticipate present/future torque demands based on various inputs such as but not limited to road profile in front, curves, gradient, etc., and present load/comfort requirements. For example, even though weather outside is not so hot there may be demand to further cool interiors, or lights may be ON even in day time. The MPC based logic can take judicious decision to satisfy all logical requirements ensuring that there is no unnecessary wastage of energy. Further, to maximize energy storage in battery, the regeneration can be maximized by choice of operating points (e.g., braking torque, speed) while keeping temperatures within safe limits.

[00136] FIG. 8 illustrates an exemplary block diagram 800 of MPC based supervisory controller for electric vehicles (EV) in accordance with an aspect of the present disclosure. As shown in FIG. 8, the block diagram represents a supervisory controller 804, motor(s)/generator(s)806, a battery pack 808, a torque coupling 810, a differential mechanism 812, wheels 814, a driving recommendation unit 816, and a brake system 818.

[00137] Shown in FIG. 8 is a flow of required information from various vehicle systems along with inputs and outputs of the MPC technique based EV supervisory controller which provides optimal torque commands to achieve a minimal total energy consumption, maximal regeneration as well as good balance between comfort and tractive performance, while keeping temperatures of motor and battery within safe limits. The dotted line in the FIG. 8 depicts the suggestions/ indications given by the MPC technique based supervisory controller 804 to driver through the driving recommendation unit 816 to achieve a good balance between comfort and performance while minimizing total energy consumption.

[00138] In one implementation, as shown in FIG. 8, the supervisory controller 804 includes one or more processor(s) 822 and a processing engine(s) 824.

[00139] The one or more processor(s) 822 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that manipulate data based on operational instructions. [00140] The processing engine(s) 824 may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 824. In example described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processing engine(s) 824 may be processor executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing engine(s) 824 may include a processing resource (for example, one or more processors), to execute such instructions. In the present example, the machine-readable storage medium may store instructions that, when executed by the processing resource, implement the processing engine(s) 824. In said example, the supervisory controller 804 may include the machine-readable storage medium storing the instructions and the processing resource to execute the instructions, or the machine-readable storage medium may be separate but accessible to supervisory controller804 and the one or more processor(s) 824. In other example, the processing engine(s) 824 may be implemented by electronic circuitry.

[00141] In an implementation, the processing engine(s) 824 comprises a vehicle parameter input receive module 826, a torque assist value generation module 828, and a torque assist value split module 830. Although the modules 826, 828, and 830 are shown as a part of the supervisory controller 804; however, these modules can be disposed outside the supervisory controller 804 and operatively connected/coupled to the supervisory controller 804, without deviating from the scope of the present disclosure.

[00142] In operation for EV, the vehicle parameter input receive module 826 receives, at the supervisory controller 804, vehicle parameter inputs 802 from throttle position sensor, brake position sensor, battery and motor voltages, temperatures and currents. In one example, the vehicle parameter inputs 802 can be selected from any or a combination of brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltages, state of charge (SOC) of battery, battery and motor temperatures, and battery and motor currents.

[00143] Upon receipt of the vehicle parameter inputs 802, the torque/torque value generation module 828 processes the received vehicle parameter inputs 802 using a model predictive control (MPC) technique to generate an optimal torque value. The generated optimal torque value is split into a tractive torque component and a load torque component by the torque assist value split module 830. The said load torque component is used for one or more loads that form part of the EV in a manner so as to minimize energy consumption. [00144] In an implementation, the supervisory controller 804 is configured to provide recommendations to the driver through driving recommendation unit 816 and/or provide input commands to motors/generators 806 to achieve the desired optimization. Further, the intended structure of some of the components of the MPC based supervisory controller 804 to achieve desired performance in EV are formulated as:

Cost function:

a. Total Energy Consumption (TEC) to be minimized. For an EV, all electrical loads are to be considered. Hence,

The first term in RHS is the motor power consumption, while the other terms are that of connected electrical loads.

b. Temperature of motor, t_m to be minimized

c. Temperature of battery, t_bto be minimized

d. Error of command torque w.r.t torque of sweet spot, to be minimized:

Here, the instantaneous torque commands are to be given such that they are as close as possible to the corresponding sweet spot torques belonging to optimal region, resulting in error minimization

e. Regeneration Energy (RE), to be maximized. That is, minimize

Drive Cycle

RE *= ^ —Vmregen(i)Tb(i)^wb(i) At

i=l

Total Cost Function=w_A TEC + w₂t_m + w₃t_b + w₄E_CST + w₅RE *

Where, wi, ... , w₅ are variable (dynamically changing) weights given to the optimization routine at each iteration from a higher level logic, based on the current operating conditions, preferences and requirements. Constraints:

f. Acceleration, a < ata ie (This ata ie is the recommended table based on current speed) g. t_m≤ t_m* (threshold temperature of motor)

h. ¾ < t_b* (threshold temperature of battery)

i. At_m < tm* (based on the current value of t_m)

j .At_b < Δ¾* (based on the current value of t_b)

k. Ti + + ΔΓ₂ + ΔΓ₃ ... . +ΔΓ_η = Tf {final torque desired)

(ΔΤι are incremental torques aligned to the best possible optimal path, as per (d), yet finally they will lead to the final required torque to be delivered)

[00145] In accordance with the present disclosure, generic implementation process 900 for MPC based supervisory controller in retrofit HEV as well pure EV cases are described with reference to FIG. 9. In FIG. 9, an optimization routine implemented in the MPC controller for the retrofit HEV as well as pure EV is described with the help of a flow chart. At block 902, the process for MPC based supervisory controller in retrofit HEV as well pure EV is initiated. At block 904, a cost function formulated by certain variables as states, error of outputs with respect to the references and outputs as its terms are realized or by A, B, C, D matrices (due to linear MPC). An objective of the cost function is to minimize energy consumption, maximize regeneration, while maintaining safe temperatures of components. At block 906, weights assigned to A, B, C, D matrices to receive weight matrices A, B, C, D. In an example, the weights are to be chosen based on the current operating conditions and several high-level requirements. At block 908, current states, references, and outputs are realized and received. At block 910, exogenous model and its input are received. At block 912, constraints for input are received. At block 914, prediction matrices are calculated. At block 916, optimization routine in accordance with the present subject matter is initialized. At block 918, input constraint from the output constraint is calculated. At block 920, the input from the optimization routine is calculated.

[00146] FIG. 10 illustrates an exemplary functional structure 1000 for implementing operation of MPC operating in supervisory controller for a plant such as EV/retrofit HEV. The functional structure 1000 comprises blocks such as input constraints 1002, an MPC 1004, a control command (first command 1006, and a plant responds 1008. In an example, the MPC 1004 further comprises a cost function 1010 and an observer/prediction 1012, for implementing the steps of the MPC process 900 depicted in FIG. 9. Generic steps to implement MP C in retrofit HEV /pure EV

Step 1 : Create control oriented plant model in state space.

Step 2: Set initial control and states.

Step 3 : Predict states over a set horizon.

Step 4: State feedback for cost function.

Step 5: Realize cost function using the required components e.g. states, set point, etc.

Step 6: Minimize cost function (using optimization routines) over the prediction horizon. Step 7: Get the control vector that minimizes the cost function over the horizon.

Step 8: Apply the first control and discard rest.

Step 9: Get state feedback.

Step 10: Repeat step 3 onwards by forward shifting the start point by one from previous iteration.

[00147] In an aspect, FIG. 11 illustrates a method 1100 to port complex MPC Code structure into simple hardware in accordance with the present disclosure. The method 1100 comprises the following steps: at block 1102, for calculation of optimal torque, reducing prediction horizon (based on operating tractive dynamic responses) to reduce memory consumed in ram, as it reduces dimension of resultant prediction matrices, the method generates an optimal motor torque assist value irrespective of parameters associated with an operational engine of the retrofitted HEV, while optimizing memory occupied by code of MPC (e.g. splitting of functions with suitable memory sizes, fitting into memory map of processor, and converting to macros for iterative calls) in ROM at step 1104, and creates a proportion of a torque share based on said optimal motor torque assist value at block 1106, for reducing fuel and/or energy consumption and emissions of the retrofitted HEV, while optimizing the RAM consumption (e.g. changed datatypes, minimizing dynamic variables in main function and alternatively sending across called macros).

[00148] In an aspect, to extend capability of the present subject matter to also address problems of minimising energy consumption and emission along with better comfort and drivability as well, additional cost functions from methods like ECMS, PEARS, SMPCL, and the like can be added to the existing MPC cost function with suitable weights. [00149] Furthermore, in order to get better optimization results, the prediction output of model can be improved by feeding known future inputs in case of repetitive routes as a result of learning algorithm based on distance mapping.

[00150] While the foregoing describes various embodiments of the present disclosure, other and further embodiments of the present disclosure may be devised without departing from the basic scope thereof. The scope of the present disclosure is determined by the claims that follow. The present disclosure is not limited to the described embodiments, versions or examples, which are included to enable a person having ordinary skill in the art to make and use the present disclosure when combined with information and knowledge available to the person having ordinary skill in the art.

ADVANTAGES OF THE INVENTION

[00151] The present disclosure achieves best possible and reliable performance, while respecting constraints such as safety for a pure EV as well as a retrofitted HEV, using MPC framework.

[00152] The MPC framework according to the present disclosure for a retrofit HEV needs no interface with the engine sensors or actuators, but results in pushing engine operation towards optimal region and savings in fuel.

[00153] The MPC framework for the case of retrofit HEV mode according to the present disclosure wherein at each instant of execution step of the controller, the optimal motor torque assist value by the supervisory controller will correspond to the best possible reduction in overall energy consumption, fuel consumption and emissions at the same time enable the system to be limited within the safety constraints. For the case of pure EV, the MPC framework in the supervisory controller will operate in EV mode to plan/arbitrate allocation of motor torque for tractive purpose (e.g., operate motor close to optimal region), HVAC load, electric loads (lights, wipers, etc.) such that the total electric energy consumption is minimal and regeneration is maximum, while maintaining safe temperature limits for motor, motor controller and battery.

[00154] The MPC framework according to the present disclosure can deal with abnormal/fault conditions as well as reduce jerks/torque pulsations for pure EV as well as retrofit HEV cases. [00155] The MPC framework in case of retrofit HEV according to the present disclosure is not dependent on the access to engine sensors and actuators throughout the entire vehicle operation to achieve the same benefits as a conventional MPC based HEV supervisor.

Claims

We Claim:

1. A model predictive control based system for a retrofit hybrid electric vehicle (HEV), said system comprising:

a vehicle parameter input receive module, which when executed by a supervisory controller that forms part of the system, receives one or more vehicle parameter inputs at the supervisory controller;

a torque assist value generation module, which when executed by the supervisory controller, processes the received one or more vehicle parameter inputs using a model predictive control (MPC) technique to generate an optimal torque assist value, said optimal torque assist value being optimally split into a first torque component that is required for a vehicle engine and a second torque component that is required for vehicle motor in a manner so as to minimize fuel consumption; and

a motor operation module, which when executed by the supervisory controller, operates a vehicle motor of the retrofit HEV using a motor torque command, sent by the supervisory controller to a motor controller of the vehicle motor, based on the second torque component.

2. The system of claim 1, wherein the MPC technique comprises a prediction mechanism that predicts dynamic operational engine variables, based on which the optimal torque assist value is generated.

3. The system of claim 2, wherein the system generates a cost function based on any or a combination of the predicted dynamic operational engine variables, vehicle speed, motor torque, motor power, motor efficiency, engine power, engine efficiency, and one or more constraints pertaining to any or a combination of motor and battery parameters the predicted dynamic operational engine variables, said cost function being used by said MPC technique to minimize the fuel consumption and overall energy consumption.

4. The system of claim 3, wherein said predicted dynamic operational engine variables are selected from any or a combination of vehicle speed, motor power and efficiency, engine power and efficiency, and one or more constraints pertaining to motor and/or battery parameters.

5. The system of claim 1, wherein the one or more vehicle parameter inputs are selected from any or a combination of brake pedal position, throttle position value, gear position, vehicle speed, battery and motor voltages, state of charge (SOC) of battery, battery and motor temperatures, and battery and motor currents.

6. The system of claim 1, wherein the MPC technique optimizes generation of the optimal motor torque assist value in abnormal conditions by using adaptive weightages and scaling factors, wherein the MPC technique assigns the adaptive weightages to one or more vehicle parameter inputs, and wherein said MPC technique enables computation of the optimal torque assist value based on a cost function factor of predicted fuel consumption.

7. The system of claim 1, wherein the supervisory controller is configured to transmit one or more recommendations to a driver of the retrofit HEV pertaining to any or a combination of optimal gear position and optimal accelerator pedal position so as to minimize the fuel consumption.

8. The system of claim 1, wherein the motor torque command is further optimally split if said retrofit HEV comprises a plurality of vehicle motors.

9. The system of claim 1, wherein said supervisory controller predicts future dynamics of the retrofit HEV based on the MPC technique.

10. The system of claim 1, wherein the supervisory controller maximizes regeneration based on the MPC technique.

11. The system of claim 1, wherein the system is operatively coupled with a post processing block that incorporates the optimal torque assist value so as to minimize j erk/torque pul sati on .

12. The system of claim 1, wherein the MPC technique is configured to predict present/future torque demands based on one or more inputs selected from any or a combination of road profile in front, road attributes, curves, and gradient.

13. The system of claim 1, wherein the MPC technique is configured to predict any or a combination of present load/comfort requirements and distance mapping.

14. The system of claim 1, further comprises a porting technique to port a complex MPC code structure into a hardware for real time control operations in retrofit HEV vehicles, wherein the porting technique comprises:

reducing a prediction horizon (based on operating tractive dynamic responses) in a MPC implementation to reduce a memory consumed;

optimizing the memory occupied by the code of MPC;

removing at least repetitive calculations of states in the system, and utilizing optimization routine(s) during the implementation of the MPC.

15. A system for an electric vehicle (EV), said system comprising:

a torque assist value generation module, which when executed by the supervisory controller, processes the received one or more vehicle parameter inputs using a model predictive control (MPC) technique to generate an optimal torque/torque value; and

a torque assist value split module, which when executed by the supervisory controller, optimally splits the optimal torque/torque value into a tractive torque component and a load torque component, said load torque component being used for one or more loads that form part of the EV in a manner so as to minimize energy consumption.

16. The system of claim 14, wherein the MPC technique comprises a prediction mechanism that further predicts future torque demands based on any or a combination of road profile in front, load requirement, and comfort requirement.

17. The system of claim 14, wherein the system generates a cost function based on any or a combination of motor power, motor efficiency, temperature of motor or battery, regeneration, error in optimal torque or load and one or more constraints pertaining to any or a combination of motor and battery parameters, acceleration, and torque requirement, said cost function being used by said MPC technique to generate the optimal torque value.