Energy management system for an electric vehicle

Abstract

An energy management system (102) for an electric vehicle (100), the energy management system comprising: a rechargeable energy storage system (104), RESS (104); an HVAC subsystem (116), comprising: a local HVAC processor (118); at least one HVAC memory (120) containing instructions that can be executed by the local HVAC processor (118) such that the local HVAC processor (118) is programmed to generate a nominal signal associated with a request for basic power input from the RESS (104); an HVAC actuator (122) that is capable of generating a target output over a predetermined period of time in response to the HVAC actuator (122) receiving the basic power input from the RESS (104); a propulsion subsystem (136), comprising: a local drive processor (138); at least one drive memory (140) that stores instructions executable by the local drive processor (138), such that the local drive processor (138) is programmed to generate a drive signal associated with a request for drive power obtained from the RESS (104); and a monitoring computer (146) comprising: a monitoring processor (148); and a monitoring memory (150) containing instructions so that the monitoring processor (148) is programmed to: to determine a value function V based on a multitude of actions U in a multitude of states S; and to select an action associated with the highest reward value (R) corresponding to the value function V; where at least one of the actions U includes an HVAC subsystem variable; wherein at least one of the states S comprises at least one of the states selected from the group consisting of the drive power to operate the drive subsystem (136), a base power input to operate the HVAC subsystem (116), a nominal reference cabin heat input setpoint determined by the local HVAC processor (118), an electric vehicle acceleration (100), a current vehicle speed, an average vehicle speed, and a calibrated estimate of the electric vehicle average speed (100); where the HVAC actuator (122) is configured to actuate an HVAC component (124) which is configured: to operate with an initial efficiency to generate an initial output when the electric vehicle (100) is in an initial state and the HVAC component (124) receives an initial power input from the RESS (104); to operate with a second efficiency to generate a second output when the electric vehicle (100) is in a second state and the HVAC component (124) receives a second power input from the RESS (104); and to generate the target output in response to the HVAC actuator (122) receiving a modulation between the first power input and the second power input over the predetermined time period, and wherein the second efficiency is higher than the first efficiency, such that an electrical power associated with the modulation between the first power input and the second power input over the predetermined time period is lower than the electrical power associated with the base power input over the predetermined time period; and wherein the selection by the monitoring processor (148) of the action associated with the highest reward value (R) includes the generation of a modulated power signal by the monitoring processor (148) in response to the monitoring processor (148) receiving the nominal signal from the local HVAC processor (118), and wherein the local HVAC processor (118) modulates between the first power input and the second power input in response to the local HVAC processor (118) receiving the modulated signal from the monitoring processor (148).

Description

Technical field

[0001] The present disclosure relates to an electric vehicle system and, in particular, to an energy management system for an electric vehicle (EV) that coordinates multiple loads to improve battery life and increase the range of the EV while simultaneously meeting the driving requirements of the driver and the comfort of the cabin. Introduction

[0002] Current operating strategies for electric vehicles do not coordinate the various load demands (propulsion, HVAC, heating, etc.) to manage their combined impact on battery usage and long-term battery health. Because multiple consumers can draw power from the battery simultaneously, battery current can spike, potentially accelerating battery aging. Furthermore, current strategies can supply power to consumers without regard to their efficiency cycles. Thus, a load operating below average efficiency can consume more electrical energy than the same load operating above average efficiency when generating constant power across different states.

[0003] DE 10 2018 211 575 A1 describes a method for operating an electrical system of a motor vehicle, wherein the electrical system comprises at least one rechargeable electric battery and one electric motor usable as a drive motor and generator, and wherein at least one control unit assigned to at least one component is used to control the components according to operating parameters describing the energy input and output of the components, wherein the operating parameters are determined as output data of a determination algorithm using and / or programmed by artificial intelligence, which determines the output data from input data describing the current operating state of the motor vehicle.

[0004] German patent DE 10 2013 211 871 A1 describes a method and a device for operating an electric vehicle. In this process, a destination is selected, a function (and thus the power consumption) of at least one vehicle component is set, and a route and distance are determined. The vehicle's range and a range buffer are calculated, and a priority list is created that describes which change to the state of the at least one vehicle component should be implemented if the calculated range is less than the calculated distance. A suggestion to change the operating state by reducing power consumption is issued, or a second destination that can be reached with the calculated range is suggested.

[0005] US 2019 / 0187269A1 describes a radar system and a method for determining true velocity, which includes: obtaining a radial velocity component corresponding to a target object based on sensor data acquired by a radar sensor on a vehicle; obtaining a true velocity magnitude of the target object from a communication protocol; and calculating a true velocity angle of the direction of the target object based on a trigonometric relationship established between the radial velocity component and the true velocity vector.

[0006] US Patent 2019 / 0064934A1 describes a method and device for detecting the lane-keeping state of vehicles in adaptive cruise control systems. An example vehicle includes a V2V communication module, an image-capturing camera, and a control unit. The control unit identifies the vehicle type of a leading vehicle based on the images and determines the speed of the leading vehicle. The control unit is also intended to send a warning to the leading vehicle via the communication module if it determines that the leading vehicle is a passenger car and its speed is below a speed setting for the activated adaptive cruise control.

[0007] German patent application DE 10 2012 216 115 A1 describes a method for controlling a vehicle. The method involves assigning a predicted driving pattern to a predicted path for the vehicle and providing a range for the vehicle using the predicted energy efficiency and an amount of energy available to the vehicle. The predicted driving pattern has an associated predicted energy efficiency. A vehicle includes a drive unit coupled to the vehicle's wheels via a transmission and a controller electronically coupled to the drive unit. The controller is configured to: (i) assign a predicted driving pattern to a predicted path for the vehicle, wherein the predicted driving pattern has a predicted energy efficiency, and (ii) provide a range for the vehicle using the predicted energy efficiency and an amount of energy available to the vehicle.

[0008] DE 198 38 248 A1 describes a method for controlling electrical consumers in an on-board electrical system, for example in a motor vehicle, in which the current state of the electrical consumers is constantly monitored and their control is carried out by at least one control unit. The publication aims to create a method for controlling electrical consumers in an on-board electrical system that ensures the best possible driving comfort under all operating conditions of the motor vehicle.The procedure includes the classification of all electrical consumers of the system into strategy groups, as well as the definition of control algorithms and priorities for the individual strategy groups, the storage of the classification and the control algorithms of the strategy groups in a memory, the monitoring of all electrical consumers by the control unit and storage of the acquired data in the memory, and the simultaneous or time-delayed control of the electrical consumers by the control unit after evaluation of the switching requirements of superior electrical consumers and release of the control according to the defined algorithms.

[0009] While existing electric vehicle systems fulfill their intended purpose, the object of the invention is to provide a new and improved energy management system that addresses these problems. Description of the invention

[0010] The invention is defined by the claims.

[0011] According to the present disclosure, a monitoring computer is provided for an energy management system of an electric vehicle. The monitoring computer comprises a monitoring processor and a monitoring memory. The monitoring memory contains instructions such that the monitoring processor is programmed to determine a value function V based on a plurality of actions U in a plurality of states S. The monitoring processor is further programmed to select an action U associated with a highest reward value at a state V corresponding to the value function V. The action U is an HVAC subsystem variable. The states S include at least one from a rechargeable energy storage system (EPS).: Rechargeable Energy Storage System (RESS) power sourced to operate a propulsion subsystem, a base power input sourced from the RESS to operate an HVAC subsystem, a nominal reference cabin heat input setpoint determined by the local HVAC processor, an electric vehicle acceleration, a current vehicle speed, an average vehicle speed, and a calibrated estimate of the average vehicle speed.

[0012] The monitoring processor can also be programmed to calculate a current reward value based on a change in battery capacity loss or cabin comfort.

[0013] The average vehicle speed can be based on vehicle-to-vehicle (V2V) data or vehicle-to-infrastructure (V2X) data.

[0014] The calibrated estimate of the vehicle's average speed can be based on previous statistics for the electric vehicle or a speed limit.

[0015] The monitoring processor can also be programmed to activate an agent based on the selected action.

[0016] The agent can include the HVAC subsystem.

[0017] The monitoring processor can also be programmed to receive the state S from a drive subsystem of the electric vehicle.

[0018] According to the invention, an energy management system for an electric vehicle is provided. The system comprises a rechargeable energy storage system (RESS) and an HVAC subsystem. The HVAC subsystem includes a local HVAC processor and at least one HVAC memory. The HVAC memory contains instructions that can be executed by the local HVAC processor, such that the local HVAC processor is programmed to generate a nominal signal associated with a request for a base power input from the RESS. The HVAC subsystem further comprises an HVAC actuator capable of generating a target output over a predetermined period of time in response to the HVAC actuator receiving the base power input from the RESS. The system also includes a drive subsystem with a local drive processor.The drive subsystem further comprises at least one drive memory that stores instructions executable by the local drive processor, such that the local drive processor is programmed to generate a drive signal associated with a request for drive power drawn from the RESS. The system further comprises a monitoring computer with a monitoring processor. The monitoring computer further comprises a monitoring memory containing instructions such that the monitoring processor is programmed to determine a value function V based on a plurality of actions U in a plurality of states S. The monitoring processor is further programmed to select an action associated with the highest reward value at a state V corresponding to the value function V. The value function V is a fitted power input setpoint for the HVAC subsystem.Action U is an HVAC subsystem variable. State S includes at least one of the drive power to operate the drive subsystem, the base power input to operate the HVAC subsystem, a nominal reference cabin heat input setpoint determined by the local HVAC processor, electric vehicle acceleration, current vehicle speed, average vehicle speed, and a calibrated estimate of the vehicle's average speed. The HVAC actuator is configured to actuate an HVAC component configured to operate at an initial efficiency to produce an initial output when the electric vehicle is in an initial state and the HVAC component receives an initial power input from the RESS.The HVAC component is further configured to operate with a second efficiency to generate a second output when the electric vehicle is in a second state and the HVAC component receives a second power input from the RESS. The HVAC component is also configured to generate the target output when it detects modulation between the first and second power inputs over the predetermined time period. The second efficiency is higher than the first, so the electrical power associated with the modulation between the first and second power inputs over the predetermined time period is lower than the electrical power associated with the baseline power input over the predetermined time period.The monitoring processor's selection of the action associated with the highest reward value involves generating a modulated power signal in response to receiving the nominal signal from the local HVAC processor. The local HVAC processor then modulates between the first and second power inputs in response to receiving this modulated signal from the monitoring processor.

[0019] According to one embodiment, the monitoring processor does not control the drive subsystem.

[0020] According to one embodiment, the monitoring processor is further programmed to calculate a current reward value R based on a change in battery capacity loss or cabin comfort.

[0021] According to one embodiment, the average vehicle speed is based on vehicle-to-vehicle data (V2V data) or vehicle-to-infrastructure data (V2X data).

[0022] According to another embodiment, the calibrated estimate of the vehicle's average speed is based on previous statistics for the electric vehicle and / or a speed limit.

[0023] According to another embodiment, the monitoring processor is further programmed to activate an agent based on the selected action.

[0024] According to another embodiment, the agent includes the HVAC subsystem.

[0025] According to another embodiment, the monitoring processor is further programmed to receive the state S from a drive subsystem of the electric vehicle.

[0026] According to several aspects of the present disclosure, a method for operating a computer for an electric vehicle energy management system is provided. The computer includes a processor and memory. The method comprises determining, using the processor, a value function V based on a plurality of actions U in a plurality of states S. The method further comprises selecting, using the processor, an action associated with the highest reward value corresponding to the value function V. The action U is an HVAC subsystem variable.State S is at least one of the following: power taken from a rechargeable energy storage system (RESS) to operate a propulsion subsystem, power taken from the RESS to operate an HVAC subsystem, current vehicle speed, electric vehicle acceleration, nominal reference cabin heat input setpoint determined by a local HVAC processor, average vehicle speed, and calibrated estimate of the vehicle's average speed.

[0027] In one aspect, the procedure also involves calculating a current reward value, using the processor, based on a change in battery capacity loss or cabin comfort.

[0028] In another aspect, the value function V is calculated based on a change in battery capacity loss and / or cabin comfort according to: U=π*(S)=arg maxuV*(S,U) where V(S, U) represents a current multidimensional table V(S, U) for the control mapping, and the current multidimensional table V(S, U) is a value function with input state S to provide the corresponding action U. Additionally, arg maxuV*(S,U) An operation to select the action U associated with the highest reward value R in state S from the current multidimensional table V(S, U). Furthermore, n*(S) is a strategy that puts the operation into a representation such that the monitoring processor selects the action U associated with the highest reward value R in the current state S from the current multidimensional table V(S, U) as the final action U. The value function V is further computed according to: Vnew(S,U)←[1−α]V(S,U)+α[R(S,U)+γmaxU'V(S',U')] where the current multidimensional table V(S, U) is updated to a new multidimensional table Vnew(S, U). The current multidimensional table V(S, U) is multiplied by [1-α] and added to a term that adds an actual reward value R and a projected value from the current multidimensional table V(S, U) based on a next action U' in a next state S'. Furthermore, α is a learning rate and γ is a discount factor.

[0029] Further applications will become apparent from the description provided herein. It should be understood that the description and specific examples serve only for illustration and are not intended to limit the scope of this disclosure. Brief description of the drawings

[0030] The drawings described here are for illustrative purposes only and are not intended to limit the scope of the present disclosure in any way. Fig. Figure 1 is a schematic view of an example of an electric vehicle with an energy management system with a computer to coordinate multiple power consumers in order to increase battery life and maintain comfort in the cabin. Fig. 2 is a block diagram of the system of Fig. 1. Fig. Figure 3 is a flowchart of an example procedure for operating the system of Fig. 1. Fig. Figure 4 is a diagram of an example modulation of the drive power and the HVAC power. Detailed description

[0031] The following description is merely exemplary and is not intended to limit the present disclosure, application or use.

[0032] This disclosure describes an example of an energy management system (system) that coordinates multiple loads to achieve optimal load delivery for improved electric vehicle (EV) range and reduced battery stress. Non-limiting examples of these loads include a drive subsystem, an HVAC subsystem, a battery heating and cooling subsystem, an accessory load, and / or other suitable subsystems. The system employs a multi-layered control strategy in which an EM processor of the monitoring layer (monitoring processor) adjusts or supplements local power requests (e.g., from an HVAC processor) based on drive variables and advance information (e.g., integrated drive and HVAC subsystem delivery). The system provides real-time feedback, e.g., in the form of energy consumption, long-term battery health, and cabin comfort.The system monitors individual loads (e.g., the HVAC control system) to minimize overall energy consumption by modulating the power delivered to load requesters based on their efficiency. As described in more detail below, the system employs reinforcement learning with a reward mechanism to train the system during a design phase and then adapt to changes in real time during operation. In one example, the system acts as a preset or customer-selectable eco-load management function (e.g., eco-cooling for HVAC) that uses drive information (e.g., speed and acceleration) to manage other loads and improve the long-term battery state of health (SOH).

[0033] An example of a motor vehicle 100 with an energy management system 102 (system) is in Fig. Figure 1 shows the vehicle 100 being an electric land vehicle, e.g., a car or a truck. The system 102 comprises a rechargeable energy storage system (RESS) 104 and a plurality of loads 106, each load 106 comprising a local processor 108 and a local memory 110 for storing instructions that can be executed by the local processor 108, such that the local processor 108 is programmed to generate a nominal signal associated with a request for energy from the RESS 104. Each load 106 further includes an actuator 112 capable of actuating various components 114 to generate a set power output over a predetermined period in response to the actuator 112 receiving current from the RESS 104.The actuators 112 are implemented by circuits, chips, motors or other electronic and / or mechanical components that can actuate various subsystems of the vehicle in accordance with suitable control signals, as is known.

[0034] As described in the examples below, each component 114 is one or more hardware components 114 that can perform a mechanical or electromechanical function, such as adjusting the blower temperature or fan speed of an HVAC subsystem. The efficiency of these components 114 depends on the environmental condition S. A component 114 operating in one condition S may have an efficiency higher than its average efficiency over a given period, and the same component may have an efficiency lower than its average efficiency over the same period in a different condition. As detailed below, the system 102 supplies more power to certain components when they are operating more efficiently, in order to reduce overall power consumption.

[0035] An exemplary load is an HVAC subsystem 116, which includes a local HVAC processor 118. The HVAC subsystem 116 further includes at least one HVAC memory 120 containing instructions that can be executed by the local HVAC processor 118, such that the local HVAC processor 118 is programmed to generate a nominal signal. The nominal signal is associated with a request for a nominal HVAC power input P. HVAC.nom . associated with the RESS 104 to provide a nominal reference cabin heat input Q cabin,nom to generate or maintain the cabin at a reference cabin temperature T cabin,nom to maintain. The HVAC subsystem 116 further includes an HVAC actuator 122, which is capable of actuating various HVAC components 124 to maintain a setpoint output Q. cabin,nom or T cabin,nom to generate over a predetermined period when the HVAC actuator 122 reaches the nominal HVAC power input P HVAC.nomfrom the RESS 104. Non-restrictive examples of HVAC components 124 include an air conditioning compressor, a radiator, a radiator fan, a condenser, and a blower.

[0036] HVAC component 124 can move into a variety of states, in which it operates with a corresponding efficiency. HVAC component 124 is configured to operate at a first efficiency to produce a first output when vehicle 100 is in a first state. HVAC actuator 122 is further configured to operate at a second efficiency to produce a second output when vehicle 100 is in a second state. In the first state, vehicle 100 travels at a first vehicle speed, and in the second state, vehicle 100 travels at a second vehicle speed that is higher than the first vehicle speed, so that the second efficiency is higher than the first efficiency, and the second output is higher than the first output over an identical period.Since, as just one example, the radiator transfers heat from the coolant to the airflow through the radiator, the efficiency of the HVAC subsystem 116 can be directly proportional to the vehicle speed, e.g., if the vehicle 100 is traveling at a first speed in the first state and the vehicle 100 is traveling at a second speed that is higher than the first speed in the second state. As detailed below, the system 102 can modulate the power delivered to the HVAC components between the first and second HVAC power inputs over a predetermined period such that the cumulative first and second outputs have the same target output Q. cabin,nom or T cabin,nom supply which the HVAC component 124 is able to supply when it has a fixed nominal HVAC power input P HVAC.nom to the HVAC components 124 over the same period. Furthermore, the cumulative modulated power P HVAClower than the cumulative fixed nominal HVAC power P HVAC.nom over the same period, because the initial HVAC power input is lower than the nominal HVAC power input Q cabin,nom , and which is supplied to the HVAC component 124 when it operates at the lower first efficiency, and the second HVAC power input is higher than the nominal HVAC power input Q cabin,nom , and which is supplied to the HVAC component 124 when it operates at the higher second efficiency.

[0037] Another exemplary load can comprise a thermal battery cooling and heating subsystem 126 (thermal subsystem 126) which includes a local thermal processor 128. The thermal subsystem 126 further comprises at least one thermal memory 130 containing instructions that can be executed by the local thermal processor 128, such that the local thermal processor 128 is programmed to generate a nominal signal. The nominal signal is associated with a request for a nominal thermal power input from the RESS 104. The thermal subsystem 126 further comprises a thermal actuator 132, which is capable of actuating various thermal components to generate a target output over a predetermined period when the thermal actuator 132 receives the nominal thermal power input from the RESS 104.Similar to the HVAC actuator 122, the thermal actuator 132 is configured to generate the associated target output when it receives a modulation between the first and second power inputs over the predetermined time period. In one example, the thermal actuator 132 is configured to actuate thermal components 134, such as a heating resistance wire or other suitable heating elements and / or a fan. It is understood that the loads may include any combination of the HVAC subsystem, the thermal subsystem, or other suitable subsystems.

[0038] Another exemplary load is a drive subsystem 136, which is monitored by system 102 for coordinating the other loads. The drive subsystem 136 includes a local drive processor 138. The drive subsystem 136 further includes at least one drive memory 140 in which instructions are stored that can be executed by the local drive processor 138, such that the local drive processor 138 is programmed to generate a drive signal. The drive signal is associated with a request for drive power P. tracThe drive subsystem 136, which is supplied by the RESS 104, is associated with the drive subsystem 136. The drive subsystem 136 further includes a drive actuator 142, which is capable of actuating various thermal components 144, such as a motor drive unit, in order to generate a target output over a predetermined period in response to the thermal actuator 132 receiving the nominal thermal power input from the RESS 104. The system 142 does not modulate or change the power P. trac , which is delivered to subsystem 136, and without affecting the driver's needs.

[0039] System 102 further comprises a monitoring computer 146 with a monitoring processor 148 and at least one monitoring memory 150. The monitoring memory 150 comprises one or more forms of computer-readable media and stores instructions that can be executed by the monitoring computer 146 to perform various operations, including those disclosed herein. A vehicle communication module 154 enables the monitoring computer 146 to communicate with a server 156 via a network 152.

[0040] The monitoring processor 148 can be communicatively connected, e.g., via the vehicle communication module 154, to more than one local processor 108, which may be contained, for example, in electronic control units (ECUs) or the like in the vehicle 100, to monitor and / or control various vehicle components 114. In this example, the monitoring processor 148 is coupled with the local drive processor 138 to monitor drive variables. Non-restrictive examples of drive variables are the current vehicle speed, the current vehicle acceleration, and the drive power. Furthermore, the monitoring computer 148 can communicate via the vehicle communication module 154 with a navigation system that uses the global positioning system (GPS) 158.The monitoring processor 148 can, for example, request and receive location data of the vehicle 100, data on speed limits, traffic data, road conditions, and the like. The location data can be in a known format, e.g., in the form of geocoordinates (latitude and longitude coordinates).

[0041] The monitoring processor 148 is generally configured for communication with the vehicle communication module 154 via an internal wired and / or wireless network, e.g. a bus or similar in the vehicle 100, such as a Processor Area Network (CAN) or similar, and / or other wired and / or wireless mechanisms.

[0042] The monitoring processor 148 can send messages to and / or receive messages from various devices in the vehicle 100 via the vehicle communication module 154, e.g., vehicle sensors 160, actuators 112, vehicle components 114, a human-machine interface (HMI) 162, etc. The human-machine interface (HMI) 162 can be configured to put the system 102 into a standard or selectable EcoBoost mode, in which the system 102 is activated to coordinate the various loads of the vehicle 100. Alternatively or additionally, in cases where the monitoring processor comprises a large number of devices, the vehicle communication network can be used for communication between the devices, which are referred to in this disclosure as the monitoring computer 146.Furthermore, as mentioned below, various processors and / or vehicle sensors can supply 160 data to the monitoring computer 146.

[0043] The monitoring processor 148 is programmed to monitor propulsion states, including the current vehicle speed, current vehicle acceleration power, and propulsion power. In other examples, the monitoring processor 148 is coupled with vehicle sensors 160, which can comprise a variety of devices to provide data that can represent or influence propulsion states. Non-limiting examples of vehicle sensors 160 can include lidar (light detection and ranging) sensors 164, etc., located on top of the vehicle, behind the vehicle's windshield, around the vehicle, etc., and providing relative positions, sizes, and shapes of objects and / or conditions in the vehicle's environment.As a further, non-limiting example, one or more radar sensors 166, mounted on the vehicle's bumpers, can provide data on the distance velocity of objects (possibly including other vehicles), etc., relative to the vehicle's location. The vehicle sensors can also include camera sensor(s) 168, e.g., for front, side, and rear views, providing images from a field of view inside and / or outside the vehicle 100.

[0044] Furthermore, the monitoring processor 148 can be configured to communicate with devices outside the vehicle 100 via a vehicle-to-vehicle communication module 154 or an interface 162, for example, via vehicle-to-vehicle (V2V) communication 170 or wireless vehicle-to-infrastructure (V2X) communication 172 with a remote server 156 (typically via the network 152). The module 154 could contain one or more mechanisms through which the monitoring processor 148 can communicate, including any combination of wireless (e.g., cellular, wireless, satellite, microwave, and radio frequency) communication mechanisms and any network topology (or topologies if multiple communication mechanisms are used). Examples of communication via the module 154 include cellular, Bluetooth, and IEEE 802.11. Dedicated Short Range Communications (DSRC) and / or Wide Area Networks (WAN), including the Internet, which provide data communication services.

[0045] The network 152 includes one or more mechanisms by which a monitoring processor 148 can communicate with a server 156. Accordingly, the network 152 can be one or more of various wired or wireless communication mechanisms, including any desired combination of wired (e.g., cable and fiber optic) and / or wireless (e.g., cellular, wireless, satellite, microwave, and radio frequency) communication mechanisms and any desired network topology (or topologies if multiple communication mechanisms are used). Example communication networks include wireless communication networks (e.g., using Bluetooth, Bluetooth Low Energy (BLE), IEEE 802.11, vehicle-to-vehicle (V2V) such as Dedicated Short-Range Communications (DSRC), etc.), local area networks (LANs), and / or wide area networks (WANs), including the Internet, which provide data communication services.Server 156 can be a computer device, i.e., it can include one or more processors and one or more memories programmed to perform operations such as those disclosed herein. Furthermore, Server 156 can be accessed via the network 152, e.g., the Internet or another wide area network.

[0046] The monitoring processor 148 can receive and analyze data from sensors 160 essentially continuously, periodically, and / or on instruction from a server 156, etc. Furthermore, object classification or object identification techniques can be used, for example, in a monitoring processor 148 based on lidar sensor, camera sensor, etc. data, to identify an object type, e.g., a vehicle, a person, a rock, a pothole, a bicycle, a motorcycle, etc., as well as physical characteristics of objects that can prompt a driver to adjust drivable variables such as the current vehicle speed, the current vehicle acceleration, and the drive power.

[0047] The monitoring processor 148 is coupled to the local processors 108 of the loads and programmed to determine one or more load variables that must be adjusted or supplemented by the local processors of the loads to reduce the total power drawn by the RESS 104. As described in detail below, the monitoring processor 148 is programmed to interpret the desired cabin responses associated with the variables. For example, the monitoring processor 148 can be programmed to interpret desired cabin responses, such as a nominal cabin temperature setpoint T. cabin,nom or a nominal reference cabin heat input setpoint Q cabin,nom which is set by the local HVAC processor 118. The monitoring processor 148 can set a customized cabin temperature setpoint T. cabin Determine for the local HVAC processor 118, which can be defined by equation 1: T¯cabin=T¯cabin,nom+fRL(.) where f RL (.) represents a correction or adjustment by the monitoring processor 148, and T cabin stands for the cabin temperature setpoint adjusted by the monitoring processor 148.

[0048] In another example, the monitoring processor 148 can be programmed to use an adapted heat input setpoint Q. cabin to be determined for use by the local HVAC processor 118. The adjusted warm input setpoint Q cabin can be determined by equations 2 and 3: Q¯cabin,nom=m˙bcp(T¯cabin−Ts) Q¯cabin=Q¯cabin,nom+fRL(.) where ṁ b the flow to a passenger cabin through a blower, c p T stands for a specific heat constant. cabin represents the adjusted cabin temperature setpoint, Q cabin,nomrepresents a nominal reference cabin heat input setpoint, which was first determined by the local HVAC processor 118, f RL (.) represents a correction or adjustment by the monitoring processor 148, and Q cabin This represents an adjusted heat input setpoint for the local HVAC processor 118. It is conceivable that the monitoring processor uses other equations to calculate the adjusted cabin temperature setpoint T. cabin , of the adjusted heat input setpoint Q cabin or implement other HVAC variables.

[0049] The monitoring processor 148 is programmed to determine the effects of the current operation on the state of health (SOH) of the battery, which can be defined by equations 4, 5 and 6: Qloss(%)=100Qb,nom−Qb(Ah)Qb,nom Qloss=g(Tb,ΔSoC)⋅(Ah)n ΔQloss=∂Qloss∂Ah⋅ΔAh=n⋅g⋅Ahn−1⋅ΔAh where Q b,nomQ represents the nominal battery capacity or total charge that a fresh RESS 104 can accept. b (Ah) represents the battery capacity at a specific point in time, Q loss (%) represents a percentage capacity loss of the battery, g represents a regression or a function with its inputs in this formula (which was calibrated during a design or training phase), T b stands for battery temperature, ΔSoC represents a change in the battery's state of charge, (Ah) n ΔAh stands for ampere-hours, defined as throughput (total battery consumption by integrating the battery current), ΔAh represents a change in throughput, n is a calibration function, and ΔQ lossThis represents an incremental capacity loss of the battery due to a current process. It is conceivable that the monitoring processor implements other equations for calculating the battery's capacity loss or other variables that affect the battery's state of health (SOH) value.

[0050] The monitoring processor 148 is programmed to determine an effect of the current operation on cabin comfort, which can be defined by equations 7, 8 and 9: Tcabin,error=∑(Tcabin−Tcabin,target)2 EHTerror=∑(EHT−EHTtarget)2 PMV∈(−0.5,0.5)Comfort range where T cabin represents the current cabin temperature, T cabin,target T represents the adjusted cabin temperature determined by the monitoring processor 148. cabin , errorrepresents an error in cabin temperature between the actual cabin temperature and the adjusted cabin temperature; EHT represents a nominal equivalent homogeneous temperature (an HVAC-related variable to combine additional measurements, such as humidity, to quantify a temperature perceived by passengers) and is first determined by the local processor 108; EHT target represents an adjusted EHT setpoint determined by the monitoring processor 148; EHT errorThis represents an error in the equivalent homogeneous temperature between the measured EHT and the adjusted EHT; and PMV represents a predicted mean rating or predicted percentage dissatisfaction, which is an index predicting an average climate as rated by the passenger. It is conceivable that the monitoring processor implements other equations to calculate the error in cabin temperature between the adjusted cabin temperature and the measured cabin temperature, an error in EHT between the measured EHT and the nominal EHT, the PMV, or other cabin comfort variables.

[0051] The monitoring processor 148 is further programmed to supplement or adjust a modulated load to generate a target output with a reduced battery state of health (SOH) value, for example, by drawing the minimum amount of electrical energy from the RESS 104. The modulated load can be the HVAC subsystem 116, the thermal subsystem 126, or other suitable vehicle subsystems. The system 102 does not modulate the powertrain subsystem 136 to avoid interfering with driver input.

[0052] In one example, the monitoring processor 148 implements a physically based model for extending the modulated load based on a preview of drive states, derived, for example, from V2V communications 170, V2X communications 172, and other network communications. Continuing from the previous example, the monitoring processor 148 can be programmed to extend the local HVAC processor 118 by providing a customized heat input setpoint Q. cabin within a fixed time window T w establishes that can be defined by equations 10 and 11: Qcabin=Qcabin,nom+g∗(vspeed−vspeedave,pre) ∑T_wQcabin=∑T_wQcabin,nominal since ∑T_wvspeed=vspeedave,pre∗Tw where Q cabin,nomfor the nominal reference cabin heat input setpoint, which was first determined by the local HVAC processor; g represents a calibration parameter to adjust the scope of adjustment by the monitoring processor 148, v speed represents the current vehicle speed; v speedave,pre for an average vehicle speed over a future time window of length T w (Preview window) is positioned and can be approached with the sensors 160, the V2V communication 170, the V2X communication 172, or another preview device; and Q cabin represents the adjusted cabin temperature setpoint determined by the monitoring processor 148 for the local HVAC processor 118.

[0053] In another example, the monitoring processor 148 implements a different physically based model for extending the modulated load, based, for example, on previous statistics of the host vehicle or prescribed speed limits. More specifically, the monitoring processor can be programmed to extend the local HVAC processor 118 by providing a customized heat input setpoint Q. cabin within a variable time window T w establishes that can be defined by equations 12 and 13: Qcabin=Qcabin,nom+g∗(vspeed−vcal) Adjust Tw during operation so that ∑T_w g∗(vcal−vspeed)=0 where v cal represents a calibrated estimate of the vehicle's average speed (using historical statistics, speed limits, and the like). Based on the average speed and the calibrated average speed of the vehicle vcal The monitoring processor 148 implements equations 12 and 13 in real time to measure a time interval T. w to determine in which the adjusted heat input Q cabin equal to the original cumulative heat input Q cabin,nom is the one requested by the local HVAC processor 118.

[0054] For each of the exemplary modulated loads, the monitoring processor 148 increases the nominal setpoint when the vehicle speed v speed above an average vehicle speed v speedave,pre or v cal lies, and the monitoring processor 148 lowers the nominal setpoint when the vehicle speed v speed below the average vehicle speed v speedave,pre or v calThis means that component 114 receives more power when it operates more efficiently. It is conceivable that the monitoring processor 148 uses other equations to determine the adjusted cabin temperature setpoint T. cabin , of the adjusted cabin temperature setpoint Q cabin , or other HVAC variables, implemented.

[0055] The Monitoring Processor 148 is further configured to implement a data-driven model for extending the modulated load. During a training mode in the design phase, the Monitoring Processor 148 implements reinforcement learning (RL) to build a lookup table or determine a control mapping V(S, U) through iterative learning with reward feedback. Later, during field deployment, the Monitoring Processor 148 uses the same structure to adjust the learning control onboard and in real time with reward feedback.

[0056] Reinforcement learning (RL) is a form of goal-directed machine learning. For example, an agent can learn from interacting with its environment without relying on explicit monitoring and / or complete models of the environment. RL is a framework that models the interaction between the learning agent and its environment in terms of states S, actions U, and rewards R. At each time step, an agent receives a state S, chooses an action U based on a strategy, receives a scalar reward, and moves to the next state S'. The states S, S' can be based on one or more sensor inputs, such as sensors 160, V2V communication 170, VZX communication 172, which are indicative of environmental data. The agent's goal is to maximize an expected cumulative reward R.The agent can receive a positive scalar reward for a positive action U and a negative scalar reward for a negative action U. The agent thus "learns" by attempting to maximize the expected cumulative reward R. Although the agent is described here in the context of a vehicle, it can encompass any suitable reinforcement learning agent. In this context, the monitoring processor 148 can be referred to as the agent. While the present example of the monitoring processor 148 implements a multidimensional table, other examples of the monitoring computer 146 can be configured to implement a reinforcement learning process based on a deep neural network.

[0057] With reference to Fig. 2 the monitoring processor 148 is further programmed to determine the control mapping V(S,U) and the learning rule using equations 14, 15, 16 and 17: U=MAP(S,R) R=−(∑Qloss+α|ΔT|) U=π*(S)=arg maxu V*(S,U) Vnew(S,U)←[1−α]V(S,U)+α[R(S,U)+γmaxU' V(S',U')] where S stands for the states, such as the drive variables (e.g., v). speed , v accel , P trac etc.) and for the nominal reference setpoints (e.g. Q cabin,nom , T cabin,nom , P HVAC , etc.); U represents actions, such as adjusted setpoints for the load expansion variable (e.g., , T). cabin or Q cabin ); R represents the rewards, such as real-time feedback in the form of battery aging (e.g., capacity loss) and cabin comfort; ΔQ loss represents the incremental change in battery capacity loss during current operation; α|ΔT| represents the difference between target and actual cabin temperature to measure cabin comfort; arg maxu V*(S,U) Indicates the process of using the V(S, U) table in the controller and selecting an action U with the highest V(S, U) value in the table for the currently measured state S; π*(S) indicates the policy that translates the operation into a representation; V (S, U) represents a current multidimensional table V(S, U) with the states S and actions U and outputs an action from U that captures the value of a specific U in state S; α is the learning rate; γ is a discount factor; α[R(S,U)+γ maxU' V(S',U')] represents a projected value from the current V table at the next S and U, denoted by S' and U'; and Vnew(S, U) represents an updated multidimensional table of states S and actions U.

[0058] Overall, the monitoring processor 148 uses equations 14 to 17 to update the current V(S, U) table during learning based on its performance, which is measured by the reward value R. In other words, the reward value R is a performance measure stored in the V(S, U) table. After the training phase, the final V(S, U) table is entered into the embedded controller and used to determine the action U. The same learning rule can also be reused in the controller at a reduced learning rate to gradually adjust V(S, U) to field variations during real-world vehicle operation.

[0059] More precisely, the monitoring processor 148 is further programmed to receive the state S from the drive subsystem 136 of the electric vehicle 100. The state S includes at least one of the drive power P tracfor the operation of the drive subsystem 136, a nominal reference cabin heat input Q cabin,nom which is determined by the local HVAC processor 118, the basic power input P HVAC the one with the nominal reference cabin heat input Q cabin,nom is associated with an acceleration of the electric vehicle 100, a current vehicle speed v speed , an average vehicle speed v speedave,pre and a calibrated estimate of the vehicle's average speed v cal One or more of these states can be based on data from at least one of the sensors 160, the V2V communication 170, and the V2X communication 172. The calibrated estimate of the vehicle's average speed v cal is based on at least one past statistic for the electric vehicle and a speed limit.

[0060] The monitoring processor 148 is further programmed to determine a value function V(S, U) based on a multitude of actions U in a multitude of states S. The monitoring processor 148 is further programmed to determine a current reward value R based on a change in battery capacity loss ΔQ. loss and / or to calculate cabin comfort α|ΔT|. In this example, the current reward value R can be incrementally increased as the change in battery capacity loss decreases ΔQ. loss and / or a reduction in the difference between target and actual cabin temperature α|ΔT| during current operation.

[0061] The monitoring processor 148 is further programmed to select an action U associated with the highest reward value R, which corresponds to the current state S and the value function V(S, U). In this example, the monitoring processor 148 generates the modulated power signal in conjunction with the request to modulate between the first power input and the second power input in response to the monitoring processor 148 receiving the nominal signal from the local HVAC processor 118, and the local HVAC processor 118 modulates between the first power input and the second power input in response to the local HVAC processor 118 receiving the modulated signal from the monitoring processor 148.

[0062] The monitoring processor 148 is further programmed to actuate an agent based on the selected action. Continuing the previous example, the agent can be the HVAC subsystem 116. The monitoring processor 148 does not control the drive subsystem, so the drive response is not modified by the system 102, and the system 102 does not take any speed or torque shaping into account.

[0063] Now, with reference to Fig. 3 is an example of a method 200 for operating the monitoring computer 146 for the energy management system 102 of the in Fig. The electric vehicle 100 is shown in Figure 1. The procedure 200 begins in block 202 with the monitoring processor 148 learning or determining the value function V(S, U) based on a multitude of actions U in a multitude of states S, for example, during a training or design phase. In this example, the action U is an HVAC subsystem variable, and the states S include at least one from a power P. trac , which is drawn from a rechargeable energy storage system (RESS) to operate a propulsion subsystem, at a current vehicle speed v speed , an average vehicle speed v speedave,pre , a calibrated estimate of the vehicle's average speed v cal , an acceleration of the electric vehicle, a nominal reference cabin heat input setpoint Q cabin,nom , which is determined by a local HVAC processor 118, and a power P HVACwhich is taken from the RESS to operate the HVAC subsystem 116. During this learning phase, the multidimensional table V(S, U) is created from initial values, e.g., empirically determined data, and tested with several profiles (e.g., states S for different driving cycles and HVAC operations), and the states S evolve over time at each time step. The monitoring processor uses equations 14 to 17 to determine the action U for the current state S.

[0064] In block 204, the monitoring processor 148 calculates at least one reward value R according to equation 15, based on a change in battery capacity loss ΔQ. loss and / or cabin comfort α|ΔT|. During the learning phase, the monitoring processor 148 measures the next state S' and calculates the reward value R.

[0065] In block 206, the monitoring processor 148 computes at least one updated value function Vnew(S, U) based on equations 13 to 17. During the learning phase, the monitoring processor 148 updates the multidimensional table V(S, U) based on blocks 202 and 204 and the example learning rule. In this way, the multidimensional table V(S, U) is updated incrementally so that the corresponding reward R for each possible action U is learned, and the monitoring processor 148 can choose the best action U with the highest reward R for a given state S. Once the test profiles are consumed during this learning step, the updated multidimensional table Vnew(S, U) is finalized and fed into the final embedded processor.

[0066] In block 208, the monitoring processor 148 selects an action U corresponding to the value function V(S, U) with the highest reward value R. In these examples, the action U is a modified HVAC setpoint sent by the monitoring processor 148 to the local processor 108 of one or more subsystems managed by the system 102, so that the system 102 can modulate the power delivered to the appropriate subsystem to reduce the total power drawn from the RESS 104 while maintaining the same output from each subsystem.

[0067] During actual vehicle operation, the pre-calibrated value function V(S,U) serves as a mapping between an arbitrary state S (which is the input for the monitoring processor 148, e.g., current measurements) and the action U, which is the output applied to the subsystem as the final control. Furthermore, the same learning rule is used to further adapt the value function V(S,U) with the actual reward value R obtained. This additional learning further refines the value function V(S,U), acting as a real-time learning algorithm. Overall, this mechanism, along with the selection of the reward R (which takes into account the effects of battery aging and cabin comfort), ensures that the mapping is optimal with respect to the chosen reward R.

[0068] With reference to Fig. In certain applications, action U (4) is an addition to the cabin temperature setpoint T. cabin,nomor the nominal reference cabin heat input setpoint Q cabin,nom Taking into account a variable U in state S (e.g., current drive requirement P) trac ), the learned action U has the effect that the monitoring processor 148 gradually increases the cabin temperature setpoint when the drive power P trac is above a predetermined threshold, so that the required HVAC power is instantly below the average HVAC power demand. Likewise, if there is a sudden drop in drive power P trac When this occurs, the monitoring processor 148 gradually lowers the cabin temperature setpoint if the drive power P trac is below a predetermined threshold, so that the required HVAC power P HVAC currently exceeding an average HVAC power demand P HVAC lies. Since the adjustments to the HVAC power requirement P HVACSince the adjustments to the HVAC power demand also involve changes in the drive power P, the average temperature in the cabin generally remains unchanged. trac To counteract this, the battery current from the RESS 104 is dampened or flattened.

[0069] In general, the computer systems and / or devices described can use any number of computer operating systems, including, but not limited to, versions and / or variants of the Ford Sync® application, AppLink / Smart Device Link middleware, the Microsoft Automotive® operating system, the Microsoft Windows® operating system, the Unix operating system (e.g., the Solaris® operating system distributed by Oracle Corporation in Redwood Shores, California), the AIX UNIX operating system distributed by International Business Machines in Armonk, New York, the Linux operating system, the Mac OSX operating system and the iOS operating system distributed by Apple Inc. in Cupertino, California, the BlackBerry OS distributed by Blackberry, Ltd. in Waterloo, Canada, and the Android operating system distributed by Google, Inc.and the Open Handset Alliance, or the QNX® CAR infotainment platform offered by QNX Software Systems. Examples of data processing equipment include, but are not limited to, an in-vehicle computer, a computer workstation, a server, a desktop, notebook, laptop, or handheld computer, or any other computer system and / or device.

[0070] Computers and computing devices generally contain computer-executable instructions, which can be executed by one or more computing devices, such as those mentioned above. Computer-executable instructions can be compiled or interpreted from computer programs created using a wide variety of programming languages ​​and / or technologies, including, but not limited to, and either alone or in combination, Java™, C, C++, MATLAB, Simulink, Stateflow, Visual Basic, Java Script, Perl, HTML, TensorFlow, Pytorch, Keras, and others. Some of these applications can be compiled and run on a virtual machine, such as the Java Virtual Machine, the Dalvik virtual machine, or similar. Generally, a processor (e.g., a microprocessor) receives instructions, for example, from memory, a computer-readable medium, etc., and executes these instructions, thereby executing one or more processes, including one or more of the processes described herein. Such instructions and other data can be stored and transmitted using a variety of computer-readable media. A file in a data processing system is generally a collection of data stored on a computer-readable medium, such as a storage medium, random-access memory, etc.

[0071] Storage can include a computer-readable medium (also called a processor-readable medium), which encompasses any non-volatile (e.g., tangible) medium involved in providing data (e.g., instructions) that can be read by a computer (e.g., a computer's processor). Such a medium can take many forms, including, but not limited to, non-volatile and volatile media. Non-volatile media include, for example, optical or magnetic hard disks and other permanent storage devices. Volatile media include, for example, dynamic random-access memory (DRAM), which typically constitutes main memory. Such instructions can be transmitted over one or more transmission media, including coaxial cable, copper wire, and fiber optic cable, including the wires that form a system bus connected to a controller's processor.Common forms of computer-readable media include, for example, a floppy disk, a flexible disk, a hard disk, a magnetic tape, any other magnetic medium, a CD-ROM, a DVD, any other optical medium, punched cards, a paper tape, any other physical medium with hole patterns, a RAM, a PROM, an EPROM, a FLASH EEPROM, any other memory chip, or a cassette, or any other medium that a computer can read from.

[0072] Databases, data repositories, or other data storage devices described herein can comprise various types of mechanisms for storing, accessing, and querying different kinds of data, including a hierarchical database, a set of files in a file system, an application database in a proprietary format, a relational database management system (RDBMS), and so on. Each of these data storage devices is generally contained within a computer device that uses a computer operating system such as one of those mentioned above, and access to it is via a network in one or more of the most varied ways. A file system can be accessed from a computer operating system and can contain files stored in various formats.An RDBMS generally uses the Structured Query Language (SQL) in addition to a language for creating, storing, editing and executing stored procedures, such as the PL / SQL language mentioned above.

[0073] In some examples, system elements may be implemented as computer-readable instructions (e.g., software) on one or more computer devices (e.g., servers, personal computers, etc.) stored on associated computer-readable media (e.g., floppy disks, memory, etc.). A computer program product may contain such instructions stored on computer-readable media to perform the functions described herein.

[0074] With regard to the media, processes, systems, methods, heuristics, etc., described herein, it is to be understood that, although the steps of such processes, etc., have been described as proceeding in a specific, ordered sequence, such processes with the described steps may be carried out in a different sequence than that described herein. Furthermore, certain steps may be carried out simultaneously, other steps may be added, or certain steps described herein may be omitted. In other words, the descriptions of methods contained herein serve to illustrate certain embodiments and are in no way to be interpreted as limiting the claims.

Claims

[1] An energy management system (102) for an electric vehicle (100), comprising the energy management system: a rechargeable energy storage system (104), RESS (104); an HVAC subsystem (116), comprising: a local HVAC processor (118); at least one HVAC memory (120) containing instructions that can be executed by the local HVAC processor (118) such that the local HVAC processor (118) is programmed to generate a nominal signal associated with a request for basic power input from the RESS (104); an HVAC actuator (122) that is capable of generating a target output over a predetermined period of time in response to the HVAC actuator (122) receiving the basic power input from the RESS (104); a propulsion subsystem (136), comprising: a local drive processor (138); at least one drive memory (140) that stores instructions executable by the local drive processor (138), such that the local drive processor (138) is programmed to generate a drive signal associated with a request for drive power obtained from the RESS (104); and a monitoring computer (146) comprising: a monitoring processor (148); and a monitoring memory (150) containing instructions so that the monitoring processor (148) is programmed to: to determine a value function V based on a multitude of actions U in a multitude of states S; and to select an action associated with the highest reward value (R) corresponding to the value function V; where at least one of the actions U includes an HVAC subsystem variable; wherein at least one of the states S comprises at least one of the states selected from the group consisting of the drive power to operate the drive subsystem (136), a base power input to operate the HVAC subsystem (116), a nominal reference cabin heat input setpoint determined by the local HVAC processor (118), an electric vehicle acceleration (100), a current vehicle speed, an average vehicle speed, and a calibrated estimate of the electric vehicle average speed (100); where the HVAC actuator (122) is configured to actuate an HVAC component (124) which is configured: to operate with an initial efficiency to generate an initial output when the electric vehicle (100) is in an initial state and the HVAC component (124) receives an initial power input from the RESS (104); to operate with a second efficiency to generate a second output when the electric vehicle (100) is in a second state and the HVAC component (124) receives a second power input from the RESS (104); and to generate the target output in response to the HVAC actuator (122) receiving a modulation between the first power input and the second power input over the predetermined time period, and wherein the second efficiency is higher than the first efficiency, such that an electrical power associated with the modulation between the first power input and the second power input over the predetermined time period is lower than the electrical power associated with the base power input over the predetermined time period; and wherein the selection by the monitoring processor (148) of the action associated with the highest reward value (R) includes the generation of a modulated power signal by the monitoring processor (148) in response to the monitoring processor (148) receiving the nominal signal from the local HVAC processor (118), and wherein the local HVAC processor (118) modulates between the first power input and the second power input in response to the local HVAC processor (118) receiving the modulated signal from the monitoring processor (148). [2] Energy management system (102) according to claim 1, wherein the monitoring processor (148) is further programmed to calculate a current reward value based on a change in at least one of a battery capacity loss and a cabin comfort. [3] Energy management system (102) according to claim 2, wherein the average speed of the electric vehicle (100) is based on vehicle-to-vehicle data, V2V data, and / or vehicle-to-infrastructure data, V2X data. [4] Energy management system (102) according to claim 3, wherein the calibrated estimate of the average speed of the vehicle (100) is based on past statistics for the electric vehicle (100) and / or a speed limit. [5] Energy management system (102) according to claim 4, wherein the monitoring processor (148) is further programmed to activate an agent based on the selected action. [6] Energy management system (102) according to claim 5, wherein the agent comprises the HVAC subsystem (116). [7] Energy management system (102) according to claim 6, wherein the monitoring processor (148) is further programmed to receive one of the states S from the drive subsystem (136) of the electric vehicle (100).

Images