A fuel cell hybrid power truck layered energy management method and system

Abstract

The application discloses a fuel cell hybrid heavy truck layered energy management method and system, the method comprising: obtaining the whole vehicle demand power of an electric vehicle; considering the performance degradation of multiple fuel cell stacks and the performance degradation of a power battery, constructing a fuel cell hybrid heavy truck energy consumption cost minimization function; converting the fuel cell hybrid heavy truck energy consumption cost minimization function and solving it through a deep deterministic policy gradient algorithm to obtain a power distribution scheme of the multiple fuel cell stacks and a power distribution scheme of the power battery; distributing power through a virtual droop control method to complete layered energy management of the fuel cell hybrid heavy truck. The embodiment of the application can perform layered energy management on the fuel cell hybrid heavy truck, realize reasonable distribution of power among multiple energy sources of the fuel cell hybrid heavy truck, and can be widely applied to the technical field of fuel cell electric vehicle energy management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fuel cell electric vehicle energy management, and particularly relates to a fuel cell hybrid heavy truck layered energy management method and system. BACKGROUND

[0002] The power system of a fuel cell hybrid heavy truck is usually composed of a multi-stack fuel cell system and a power battery. The power distribution between the multi-stack fuel cell and the power battery and between each stack has an important influence on the vehicle economy and the service life of the energy source. A reasonable energy management strategy can greatly help improve the fuel economy of the vehicle and prolong the service life of the energy source.

[0003] Related fuel cell hybrid vehicle energy management strategies include centralized control and decentralized control. The centralized control is based on a central controller, which collects the demand power and the state information of each energy source, calculates the power and current instructions of each energy source, and issues the corresponding power or current instructions to the local controllers of each energy source. Since the centralized control involves centralized communication, communication delay problems inevitably exist in actual work, resulting in poor power distribution performance. In addition, the centralized processing of signals may affect the entire system in the event of a single point failure, so it is difficult to meet the requirements of high reliability and strong flexibility of the system. The decentralized control has its own independent local controller in each unit of the system. Each unit can operate independently and stably according to the local controller. However, the work of each energy source in the decentralized control architecture is independent, and there is a lack of information interaction, making it difficult to achieve optimal scheduling of the system.

[0004] In summary, the technical problems in the related art need to be improved. SUMMARY

[0005] The main purpose of the embodiments of the present application is to provide a fuel cell hybrid heavy truck layered energy management method and system, which can perform layered energy management on the fuel cell hybrid heavy truck and realize reasonable power distribution between multiple energy sources of the fuel cell hybrid heavy truck.

[0006] To achieve the above purpose, one aspect of the embodiments of the present application provides a fuel cell hybrid heavy truck layered energy management method, which comprises the following steps:

[0007] A dynamics model of the electric vehicle is constructed to obtain the whole vehicle demand power of the electric vehicle;

[0008] The performance degradation of the multi-stack fuel cell and the performance degradation of the power battery are considered, and a fuel cell hybrid heavy truck energy consumption cost minimization function is constructed in combination with the whole vehicle demand power of the electric vehicle;

[0009] The fuel cell hybrid heavy truck energy consumption cost minimization function is converted and solved by a deep deterministic policy gradient algorithm to obtain the power distribution scheme of the multiple fuel cell stacks and the power distribution scheme of the power battery;

[0010] According to the power distribution scheme of the multiple fuel cell stacks and the power distribution scheme of the power battery, power distribution is performed by a virtual droop control method to complete hierarchical energy management of the fuel cell hybrid heavy truck.

[0011] In some embodiments, the expression of the dynamics model of the electric vehicle is specifically as follows:

[0012]

[0013] In the above formula, P dem represents the dynamics model of the electric vehicle, v represents the vehicle speed, f represents the rolling resistance coefficient, C D represents the air resistance coefficient, A represents the vehicle frontal area, p represents the air density, a represents the vehicle acceleration, a represents the road slope, and m represents the vehicle mass.

[0014] In some embodiments, the performance degradation of the multiple fuel cell stacks and the performance degradation of the power battery are considered to construct a fuel cell hybrid heavy truck energy consumption cost minimization function, which includes:

[0015] The voltage degradation amount of the multiple fuel cell stacks is obtained and characterized to construct a multiple fuel cell stack performance degradation model;

[0016] The battery capacity attenuation amount of the power battery is obtained and characterized to construct a power battery performance degradation model;

[0017] According to the multiple fuel cell stack performance degradation model and the power battery performance degradation model, the fuel cell hybrid heavy truck energy consumption cost is determined, which includes multiple fuel cell stack hydrogen consumption cost, power battery equivalent hydrogen consumption cost, multiple fuel cell stack performance degradation cost, and power battery performance degradation cost;

[0018] According to the whole vehicle demand power of the electric vehicle, a battery output power constraint condition is constructed;

[0019] The fuel cell hybrid heavy truck energy consumption cost is converted into a minimization problem function, and the battery output power constraint condition is introduced to construct the fuel cell hybrid heavy truck energy consumption cost minimization function.

[0020] In some embodiments, the expression of the minimization problem function of the fuel cell hybrid heavy truck energy consumption cost is specifically as follows:

[0021]

[0022] In the above formula, J represents a minimization problem function of fuel cell hybrid power truck energy consumption cost, C mfcs represents multi-stack fuel cell hydrogen consumption cost, C bat represents power battery equivalent hydrogen consumption cost, C mfcsd represents multi-stack fuel cell performance degradation cost, C batd represents power battery performance degradation cost, t represents t time, and T represents total time.

[0023] In some embodiments, the expression of the battery output power constraint condition is specifically as follows:

[0024]

[0025] In the above formula, P d3m,t represents the demand power of the whole vehicle at t time, and respectively represent the minimum power and the maximum power of the multi-stack fuel cell allowable output, and respectively represent the minimum power and the maximum power of the power battery allowable output, SOC min and SOC max represent the minimum SOC value and the maximum SOC value of the power battery, P mfcs,t represents the output power of the multi-stack fuel cell at t time, P bat,t represents the output power of the power battery at t time, and SOC represents the SOC value of the power battery.

[0026] In some embodiments, the fuel cell hybrid power truck energy consumption cost minimization function is converted and solved by a deep deterministic policy gradient algorithm to obtain the power distribution scheme of the multi-stack fuel cell and the power distribution scheme of the power battery, including:

[0027] A state quantity, an action quantity and a reward function are set, the fuel cell hybrid power truck energy consumption cost minimization function is converted to obtain a fuel cell hybrid power truck energy consumption cost Markov game problem, the state quantity includes the whole vehicle demand power of the electric vehicle, the voltage degradation of the multi-stack fuel cell, the battery capacity attenuation of the power battery and the SOC value of the power battery, the action quantity includes the output power of the multi-stack fuel cell and the output power of the power battery, and the reward function is fuel cell hybrid power truck energy consumption cost.

[0028] interact data of the fuel cell hybrid heavy truck energy consumption cost Markov game problem and the deep deterministic policy gradient algorithm, and train the deep deterministic policy gradient algorithm through the interact data to obtain a trained deep deterministic policy gradient algorithm;

[0029] obtain state quantity of the electric vehicle at the current time and input the state quantity to the trained deep deterministic policy gradient algorithm to obtain action quantity of the electric vehicle at the current time, the action quantity of the electric vehicle at the current time including the output power of the multiple stacks of fuel cells and the output power of the power battery;

[0030] cycle the step of obtaining state quantity of the electric vehicle at the current time and inputting the state quantity to the trained deep deterministic policy gradient algorithm to obtain action quantity of the electric vehicle at the current time until the electric vehicle stops running, and output the power distribution scheme of the multiple stacks of fuel cells and the power distribution scheme of the power battery.

[0031] In some embodiments, the expression of the fuel cell hybrid heavy truck energy consumption cost Markov game problem is specifically as follows:

[0032] s=(P dem ,V mfcsd ,Q loss ,SOC)

[0033] a=(P mfcs ,P bat )

[0034] r=C mfcs +C bat +C batd +C mfcsd

[0035] In the above formula, s represents state quantity, a represents action quantity, r represents a reward function, P dem represents the whole vehicle demand power of the electric vehicle, V mfcsd represents the voltage decay of the multiple stacks of fuel cells, Q loss represents the battery capacity attenuation of the power battery, SOC represents the SOC value of the power battery, P mfcs represents the output power of the multiple stacks of fuel cells, P bat represents the output power of the power battery, C mfcs represents the hydrogen consumption cost of the multiple stacks of fuel cells, C bat represents the equivalent hydrogen consumption cost of the power battery, C mfcsd represents the performance degradation cost of the multiple stacks of fuel cells, C batd represents the performance degradation cost of the power battery.

[0036] In some embodiments, the power distribution according to the power distribution scheme of the multiple stacks of fuel cells and the power distribution scheme of the power battery is performed by a virtual droop control method, and fuel cell hybrid heavy truck layered energy management is completed, including:

[0037] The multiple stacks of fuel cells output power is obtained based on the power distribution scheme of the multiple stacks of fuel cells, and the multiple stacks of fuel cells output power variation is calculated;

[0038] Based on the multiple stacks of fuel cells output power variation, in combination with the voltage recession amount of the multiple stacks of fuel cells, the output power variation of a plurality of stacks is determined;

[0039] The output power reference value of a plurality of stacks is determined according to the output power variation of a plurality of stacks, and the virtual droop resistance value is calculated;

[0040] The power distribution of a plurality of stacks is obtained by current adjustment of the virtual droop resistance value through the virtual droop control method;

[0041] In combination with the power distribution of a plurality of stacks and the power distribution scheme of the power battery, the layered energy management of the fuel cell hybrid heavy truck is completed.

[0042] In some embodiments, the calculation expression of the virtual droop resistance value is specifically as follows:

[0043]

[0044] In the above formula, V dci represents the voltage value after the output voltage of the i th DC / DC converter passes through the virtual resistance, V * dci represents the voltage value after the output voltage of the i th DC / DC converter passes through the virtual resistance, R Li represents the i th stack line resistance value, R di represents the virtual resistance value added to the output end of the i th DC / DC converter, I dci represents the current at the output end of the i th DC / DC converter, wherein i=1, 2, …, n, V bus represents an electric vehicle.

[0045] To achieve the above-mentioned purpose, another aspect of the embodiment of the present application proposes a fuel cell hybrid heavy truck layered energy management system, which comprises:

[0046] The first module is used for constructing the dynamics model of the electric vehicle, and obtaining the whole vehicle demand power of the electric vehicle;

[0047] a second module configured to consider performance degradation of the multiple stacks of fuel cells and performance degradation of the power battery, combine a whole vehicle demand power of the electric vehicle, and construct a fuel cell hybrid heavy truck energy consumption cost minimization function;

[0048] a third module configured to convert and solve the fuel cell hybrid heavy truck energy consumption cost minimization function by using a deep deterministic policy gradient algorithm to obtain a power distribution scheme of the multiple stacks of fuel cells and a power distribution scheme of the power battery;

[0049] a fourth module configured to perform power distribution by using a virtual droop control method according to the power distribution scheme of the multiple stacks of fuel cells and the power distribution scheme of the power battery, and complete fuel cell hybrid heavy truck hierarchical energy management.

[0050] The embodiments of the present application at least have the following beneficial effects: the present application provides a fuel cell hybrid heavy truck hierarchical energy management method and system, the scheme obtains a whole vehicle demand power of an electric vehicle, considers performance degradation of multiple stacks of fuel cells and performance degradation of a power battery, constructs a fuel cell hybrid heavy truck energy consumption cost minimization function, considers performance degradation of energy sources in the process of power distribution, improves the service life of each energy source, and further improves the economy of the fuel cell hybrid heavy truck. The scheme is solved by using a deep deterministic policy gradient algorithm, and power distribution is performed by using a virtual droop control method to complete fuel cell hybrid heavy truck hierarchical energy management, and finally realizes reasonable power distribution among multiple energy sources of the fuel cell hybrid heavy truck. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 FIG. 1 is a flowchart of a fuel cell hybrid heavy truck hierarchical energy management method provided by an embodiment of the present application;

[0052] Figure 2 FIG. 2 is a structural schematic diagram of a fuel cell hybrid heavy truck hierarchical energy management system provided by an embodiment of the present application;

[0053] Figure 3 FIG. 3 is a principle schematic diagram of a multiple stack equivalent circuit provided by an embodiment of the present application;

[0054] Figure 4 FIG. 4 is a framework schematic diagram of fuel cell hybrid heavy truck hierarchical energy management provided by an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application. When the following description refers to the accompanying drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with embodiments of the present application. They are only examples of systems and methods consistent with some aspects of the embodiments of the present application as detailed in the appended claims.

[0056] It can be understood that the terms "first", "second", and the like used in the present application can be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "when" or "when" or "in response to determining".

[0057] The terms "at least one", "multiple", "each", "any" and the like used in the present application include one, two or more than two, multiple includes two or more than two, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as understood by those skilled in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0059] Referring to Figure 1 and Figure 4 , Figure 1 A flowchart of a fuel cell hybrid power truck layered energy management method provided by an embodiment of the present application is shown in Figure 1 The method comprises the following steps:

[0060] S100, constructing a dynamics model of an electric vehicle to obtain a whole vehicle demand power of the electric vehicle;

[0061] It should be noted that in some specific embodiments, the whole vehicle demand power P dem is obtained through a vehicle dynamics model, and the expression of the dynamics model of the electric vehicle is specifically as follows:

[0062]

[0063] In the above formula, Pdem represents a dynamics model of an electric vehicle, v represents a vehicle speed, f represents a rolling resistance coefficient, C D represents an air resistance coefficient, A represents a vehicle frontal area, p represents an air density, a represents a vehicle acceleration, a represents a road slope, and m represents a vehicle mass.

[0064] S200, considering the performance degradation of the multiple stacks of fuel cells and the performance degradation of the power battery, combining the whole vehicle demand power of the electric vehicle, a fuel cell hybrid heavy truck energy consumption cost minimization function is constructed;

[0065] It should be noted that in some embodiments, step S200 can include steps S210 to S250:

[0066] S210, the voltage degradation amount of the multiple stacks of fuel cells is obtained and characterized, and a multiple stack fuel cell performance degradation model is constructed; S220, the battery capacity attenuation amount of the power battery is obtained and characterized, and a power battery performance degradation model is constructed;

[0067] In steps S210 to S220 of some embodiments, the multiple stack fuel cell performance degradation model is characterized by the voltage degradation amount of the multiple stack fuel cells, and the power battery performance degradation model is characterized by the battery capacity attenuation amount.

[0068] Wherein, the multiple stack fuel cell performance degradation model is as follows:

[0069]

[0070] In the above formula, V mfcsd represents the voltage degradation amount of the multiple stacks of fuel cells, V fcsdi represents the voltage degradation amount of the i-th stack, and n represents the number of stacks.

[0071] Wherein, the calculation of the stack voltage degradation amount is as follows:

[0072] V fcsd = V s + V i + V f + V h

[0073] In the above formula, V fcsd represents the stack voltage degradation amount, V s represents the stack voltage degradation amount under start-stop conditions, V i represents the stack voltage degradation amount under idling conditions, V f represents the stack voltage degradation amount under frequent load change conditions, V h represents the stack voltage degradation amount under heavy load conditions.

[0074] According to the voltage attenuation of the stack, the performance attenuation of each stack can be expressed by the following formula:

[0075]

[0076] In the above formula, K di represents a value representing the stack degradation, and the closer the value is to 1, the better the performance of the stack. V ideal represents the voltage value of the stack under ideal conditions.

[0077] The performance degradation model of the power battery is as follows:

[0078]

[0079] In the above formula, Q loss represents the capacity attenuation, and a and β represent the fitting coefficients of SOC, θ represents the temperature in Celsius, I c represents the charge / discharge rate, Ah represents the cumulative charge, and z represents the power index factor.

[0080] S230, according to the performance degradation model of the multi-stack fuel cell and the performance degradation model of the power battery, determine the energy consumption cost of the fuel cell hybrid heavy truck, the energy consumption cost of the fuel cell hybrid heavy truck includes the hydrogen consumption cost of the multi-stack fuel cell, the equivalent hydrogen consumption cost of the power battery, the performance degradation cost of the multi-stack fuel cell and the performance degradation cost of the power battery; S240, according to the whole vehicle demand power of the electric vehicle, construct the battery output power constraint condition; S250, convert the energy consumption cost of the fuel cell hybrid heavy truck into a minimization problem function, and introduce the battery output power constraint condition to construct the energy consumption cost minimization function of the fuel cell hybrid heavy truck.

[0081] In steps S230 to S250 of some embodiments, the cost function includes: the cost required by the hydrogen consumption of the multi-stack fuel cell, the cost required by the equivalent hydrogen consumption of the power battery, the stack performance degradation cost, and the power battery performance degradation cost.

[0082] The cost C mfcs,t of the hydrogen consumption of the multi-stack fuel cell is as follows:

[0083]

[0084] In the above formula, m mfcs represents the hydrogen consumption of the multi-stack, P mfcs represents the output power of the multi-stack, LHV represents the low heat value of hydrogen, and η fcs represents the efficiency of the stack, represents the price of hydrogen.

[0085] The equivalent hydrogen consumption m bat,t of the power battery is:

[0086]

[0087] In the above formula, P bat represents the output power of the power battery, η dis represents the discharge efficiency of the power battery, η chg represents the charge efficiency of the power battery.

[0088] The equivalent hydrogen consumption cost C bat of the power battery is:

[0089]

[0090] The performance degradation cost C mfcsd of the stack is as follows:

[0091]

[0092] In the above formula, M fcs represents the price of multiple stacks, V mfcsd_max represents the maximum voltage degradation amount allowed by the stack.

[0093] The performance degradation cost C batd of the power battery is as follows:

[0094]

[0095] In the above formula, Q loss_max represents the maximum capacity allowed to degrade by the power battery, M bat represents the price of the power battery.

[0096] Based on the above cost model, the problem of minimizing the energy consumption cost of the fuel cell hybrid heavy truck is defined as:

[0097]

[0098] In the above formula, J represents the function of the problem of minimizing the energy consumption cost of the fuel cell hybrid heavy truck, C mfcs represents the hydrogen consumption cost of multiple stacks of fuel cells, C bat represents the equivalent hydrogen consumption cost of the power battery, C mfcsd represents the performance degradation cost of multiple stacks of fuel cells, C batd represents the performance degradation cost of the power battery, t represents the time t, and T represents the total time.

[0099] In this optimization problem, the multiple stack output power, the power battery output power, and the battery SOC should satisfy the following constraint conditions:

[0100]

[0101] In the above formula, P dem,tPreq(t) represents the demand power of the whole vehicle at time t, and respectively represent the minimum power and the maximum power of the multi-stack fuel cell allowed output, and respectively represent the minimum power and the maximum power of the power battery allowed output, SOC min and SOC max represent the minimum SOC value and the maximum SOC value of the power battery allowed, P mfcs,t Pfuel(t) represents the output power of the multi-stack fuel cell at time t, bat,t Pbat(t) represents the output power of the power battery at time t, SOC represents the SOC value of the power battery.

[0102] S300, the fuel cell hybrid heavy truck energy consumption cost minimization function is converted and solved by deep deterministic policy gradient algorithm, to obtain the power distribution scheme of the multi-stack fuel cell and the power distribution scheme of the power battery;

[0103] It should be noted that in some embodiments, step S300 can include: S310, setting state quantity, action quantity and reward function, converting the fuel cell hybrid heavy truck energy consumption cost minimization function to obtain the fuel cell hybrid heavy truck energy consumption cost Markov game problem, the state quantity includes the whole vehicle demand power of the electric vehicle, the voltage degradation of the multi-stack fuel cell, the battery capacity attenuation of the power battery and the SOC value of the power battery, the action quantity includes the output power of the multi-stack fuel cell and the output power of the power battery, and the reward function is the energy consumption cost of the fuel cell hybrid heavy truck;

[0104] In step S310 of some embodiments, the energy consumption cost minimization problem is related to the power distribution between the energy source and the power distribution between the multi-stack and the power battery is completed, and the performance degradation of the multi-stack and the power battery is considered. The state quantity s is set as: the demand power P dem , the multi-stack fuel cell voltage attenuation V mfcsd , the power battery capacity attenuation Q loss and SOC. The action quantity a is set as the multi-stack fuel cell power P mfcs and the power battery power P bat , and the reward function r includes: the multi-stack fuel cell hydrogen consumption cost C mfcs , the equivalent hydrogen consumption cost C bat of the power battery, the performance degradation cost C batd of the power battery, and the performance degradation cost C mfcsd of the multi-stack fuel cell. A policy that minimizes the value function is obtained through the current state, that is, an optimal output power of the multi-stack and the power battery is obtained, and the energy consumption cost minimization problem is converted into a Markov game problem, as shown in the following formula:

[0105] s = (P dem , V mfcsd , Q loss , SOC)

[0106] a = (P mfcs , P bat )

[0107] r = C mfcs + C bat + C batd + C mfcxd

[0108] In the above formula, s represents a state quantity, a represents an action quantity, r represents a reward function, P dem represents a whole vehicle demand power of an electric vehicle, V mfcsd represents a voltage recession quantity of a multi-stack fuel cell, Q loss represents a battery capacity attenuation quantity of a power battery, and SOC represents an SOC value of the power battery, P mfcs represents an output power of the multi-stack fuel cell, P bat represents an output power of the power battery, C mfcs represents a hydrogen consumption cost of the multi-stack fuel cell, C bat represents an equivalent hydrogen consumption cost of the power battery, C mfcsd represents a performance recession cost of the multi-stack fuel cell, and C batd represents a performance recession cost of the power battery.

[0109] S320, obtain the interaction data of the fuel cell hybrid heavy truck energy consumption cost Markov game problem and the deep deterministic policy gradient algorithm, and train the deep deterministic policy gradient algorithm through the interaction data to obtain the trained deep deterministic policy gradient algorithm; S330, obtain the state quantity of the electric vehicle at the current time and input it into the trained deep deterministic policy gradient algorithm to obtain the action quantity of the electric vehicle at the current time, the action quantity of the electric vehicle at the current time including the output power of the multi-stack fuel cell and the output power of the power battery; S340, repeatedly obtain the state quantity of the electric vehicle at the current time and input it into the trained deep deterministic policy gradient algorithm to obtain the action quantity of the electric vehicle at the current time, until the electric vehicle stops running, and output the power distribution scheme of the multi-stack fuel cell and the power distribution scheme of the power battery.

[0110] In steps S320 to S340 of some embodiments, the deep deterministic policy gradient algorithm (DDPG algorithm) is used to distribute the power between the power battery and the multi-stack fuel cell of the fuel cell hybrid heavy truck, the data after the fuel cell hybrid heavy truck interacts with the DDPG algorithm is obtained, and the DDPG algorithm is trained with the data.

[0111] The current state quantity set s of the automobile is calculated by the related sensor and the related formula t = {P dem,t , V mfcsd,t , Q loss,t , SOC t} is input into the trained DDPG algorithm to obtain the current action quantity set a t = {P mfcs,t , P bat,t}, that is, the power that the multiple stacks of fuel cells and the power battery should output. The step is repeated until the automobile stops running.

[0112] S400, according to the power distribution scheme of the multiple stacks of fuel cells and the power distribution scheme of the power battery, power distribution is performed by a virtual droop control method to complete the layered energy management of the fuel cell hybrid heavy truck;

[0113] It should be noted that in some embodiments, step S400 can include: S410, obtaining the output power of the multiple stacks of fuel cells based on the power distribution scheme of the multiple stacks of fuel cells and calculating the output power variation of the multiple stacks of fuel cells; S420, determining the output power variation of a plurality of stacks based on the output power variation of the multiple stacks of fuel cells and in combination with the voltage recession of the multiple stacks of fuel cells; S430, determining the output power reference value of the plurality of stacks according to the output power variation of the plurality of stacks and calculating the virtual droop resistance value; S440, adjusting the current by the virtual droop control method to obtain the power distribution of the plurality of stacks; and S450, completing the layered energy management of the fuel cell hybrid heavy truck in combination with the power distribution of the plurality of stacks and the power distribution scheme of the power battery.

[0114] In some specific embodiments, the output power of the multiple stacks of fuel cells is obtained and the output power variation of the multiple stacks of fuel cells is calculated. According to the performance recession of each stack, the output power variation of each stack is calculated.

[0115] The output power of the multiple stacks of fuel cells is obtained by the trained DDPG model. The load demand power variation ΔP load is the output power variation of the multiple stacks, and the value is equal to the difference between the output power variation of the multiple stacks of fuel cells at the moment and the output power variation of the multiple stacks of fuel cells at the last moment. According to the performance recession K di of each stack, the output power variation of each stack is obtained, and the expression is as follows:

[0116]

[0117] In the above formula, ΔP fcsi represents the real-time output power variation of the i-th stack, where i = 1, 2, …, n, and ΔPmfcs represents the real-time output power change amount of the multiple stacks.

[0118] The output power reference value of each stack is calculated.

[0119] The power reference value of each stack is the power that each stack should output, and the calculation method is shown in the following formula:

[0120]

[0121] According to the power reference value, the virtual droop resistance value is calculated, and the current is adjusted by adjusting the virtual droop resistance, so as to realize the power distribution of each stack.

[0122] According to the equivalent circuit diagram of the multiple stacks, as shown in Figure 3 The virtual resistance value is calculated as shown in the following formula:

[0123]

[0124] In the above formula, V dci represents the voltage value of the output voltage of the i th stack after the DC / DC converter, V * dci represents the voltage value of the output voltage of the i th DC / DC converter after the virtual resistance, R Li represents the line resistance value of the i th stack, R di represents the virtual resistance value added to the output end of the i th DC / DC converter, I dci represents the current at the output end of the i th DC / DC converter, wherein i=1, 2, …, n, V bus represents an electric vehicle.

[0125] The current at the output end of the converter is represented by the real-time power that each stack should output, and the virtual resistance value can be calculated by the following formula:

[0126]

[0127] The voltage stabilization and current shunt between the DC / DC converters are realized by adjusting the virtual droop resistance, so as to complete the power distribution of each stack.

[0128] In summary, the embodiment of the application proposes a hierarchical energy management architecture of a fuel cell hybrid power system "upper layer centralized control + lower layer decentralized control", and the performance degradation of the energy source is considered in the energy distribution process, so as to realize the reasonable distribution of power among multiple energy sources.

[0129] Please refer to Figure 2The embodiment of the application further provides a layered energy management system of the fuel cell hybrid heavy truck, which can realize the layered energy management method of the fuel cell hybrid heavy truck.

[0130] The first module 201 is configured to construct a dynamics model of the electric vehicle, and obtain a whole vehicle demand power of the electric vehicle.

[0131] The second module 202 is configured to consider performance degradation of the multiple stacks of fuel cells and performance degradation of the power battery, and construct a fuel cell hybrid heavy truck energy consumption cost minimization function in combination with the whole vehicle demand power of the electric vehicle.

[0132] The third module 203 is configured to convert and process the fuel cell hybrid heavy truck energy consumption cost minimization function, and solve the fuel cell hybrid heavy truck energy consumption cost minimization function by using a deep deterministic policy gradient algorithm, so as to obtain a power distribution scheme of the multiple stacks of fuel cells and a power distribution scheme of the power battery.

[0133] The fourth module 204 is configured to distribute power by using a virtual droop control method according to the power distribution scheme of the multiple stacks of fuel cells and the power distribution scheme of the power battery, so as to complete the layered energy management of the fuel cell hybrid heavy truck.

[0134] It can be understood that the content in the method embodiment is applicable to the system embodiment, the system embodiment specifically realizes the same functions as the method embodiment, and achieves the same beneficial effects as the method embodiment.

[0135] The preferred embodiments of the application are described above with reference to the accompanying drawings, and the scope of the application is not limited by the above description. Any modification, equivalent replacement and improvement made by those skilled in the art without departing from the scope and essence of the application should be within the scope of the application.

Claims

1. A tiered energy management method for fuel cell hybrid heavy-duty trucks, characterized in that, The method includes the following steps: Construct a dynamic model of the electric vehicle to obtain the total power requirement of the electric vehicle; Considering the performance degradation of multiple fuel cells and power batteries, and combining the overall vehicle power demand of the electric vehicle, a function for minimizing the energy consumption cost of fuel cell hybrid heavy-duty trucks is constructed. The energy consumption cost minimization function of the fuel cell hybrid heavy truck is transformed and solved by a deep deterministic strategy gradient algorithm to obtain the power allocation scheme of the multi-stack fuel cell and the power battery. Based on the power distribution scheme of the multi-stack fuel cells and the power distribution scheme of the power battery, power distribution is carried out through a virtual droop control method to complete the hierarchical energy management of fuel cell hybrid heavy trucks. The process of transforming the energy consumption cost minimization function of the fuel cell hybrid heavy-duty truck and solving it using a deep deterministic strategy gradient algorithm to obtain the power allocation scheme of the multi-stack fuel cells and the power battery includes: By setting state variables, action variables, and a reward function, the function for minimizing the energy consumption cost of the fuel cell hybrid heavy-duty truck is transformed to obtain a Markov game problem for the energy consumption cost of the fuel cell hybrid heavy-duty truck. The state variables include the total power demand of the electric vehicle, the voltage decay of the multiple fuel cells, the capacity decay of the power battery, and the SOC value of the power battery. The action variables include the output power of the multiple fuel cells and the output power of the power battery. The reward function is the energy consumption cost of the fuel cell hybrid heavy-duty truck. The interaction data between the energy consumption cost Markov game problem of the fuel cell hybrid heavy truck and the deep deterministic policy gradient algorithm is obtained, and the deep deterministic policy gradient algorithm is trained through the interaction data to obtain the trained deep deterministic policy gradient algorithm. The current state of the electric vehicle is obtained and input into the trained deep deterministic policy gradient algorithm to obtain the current action of the electric vehicle. The current action of the electric vehicle includes the output power of the multi-stack fuel cell and the output power of the power battery. The process of obtaining the current state of the electric vehicle and inputting it into the trained deep deterministic policy gradient algorithm to obtain the current action of the electric vehicle continues until the electric vehicle stops running. The power allocation scheme of the multi-stack fuel cell and the power allocation scheme of the power battery are then output. The specific expression for the Markov game problem concerning the energy consumption cost of fuel cell hybrid heavy-duty trucks is as follows: In the above formula, Represents state variables. Indicates the amount of action. Represents the reward function, This indicates the total power requirement of an electric vehicle. This indicates the voltage decay of multiple fuel cell stacks. This indicates the amount of battery capacity degradation in a power battery. This indicates the SOC value of the power battery. This indicates the output power of a multi-stacking fuel cell. Indicates the output power of the power battery. This indicates the hydrogen consumption cost of multiple fuel cell stacks. This indicates the equivalent hydrogen consumption cost of the power battery. This indicates the cost of performance degradation in multi-stack fuel cells. This indicates the cost of power battery performance degradation; The step of allocating power according to the power allocation scheme of the multi-stack fuel cells and the power allocation scheme of the power battery, and completing the hierarchical energy management of fuel cell hybrid heavy-duty trucks through a virtual droop control method, includes: The output power of the multi-fuel cell stack is obtained based on the power distribution scheme of the multi-fuel cell stack, and the change in output power of the multi-fuel cell stack is calculated. Based on the output power variation of multiple fuel cells, and combined with the voltage decay of the multiple fuel cells, the output power variation of several fuel cells is determined. Based on the changes in the output power of the aforementioned stacks, determine the reference values ​​for the output power of the stacks, and calculate the virtual droop resistance value. The power distribution of several fuel cells is obtained by adjusting the current of the virtual droop resistor using the virtual droop control method. By combining the power allocation of the aforementioned fuel cell stacks with the power allocation scheme of the power battery, the hierarchical energy management of the fuel cell hybrid heavy-duty truck is completed. The specific expression for calculating the virtual droop resistance value is as follows: In the above formula, Indicates the first The voltage value of the individual fuel cell output voltage after passing through the DC / DC converter. Indicates the first The voltage value at the output terminal of a DC / DC converter after passing through a virtual resistor. Indicates the first The circuit resistance value of each fuel cell stack. Indicates the first The virtual resistor value added to the output of a DC / DC converter. Indicates the first The current at the output of each DC / DC converter, where... , It refers to electric vehicles.

2. The method according to claim 1, characterized in that, The specific expression for the dynamic model of the electric vehicle is as follows: In the above formula, Represents the dynamic model of an electric vehicle. Indicates vehicle speed. Indicates the rolling resistance coefficient. Indicates the air drag coefficient. Indicates the vehicle's frontal area. Indicates air density, Indicates vehicle acceleration. Indicates the road slope. Indicates vehicle mass.

3. The method according to claim 1, characterized in that, The aforementioned consideration of the performance degradation of multiple fuel cells and power batteries, combined with the overall vehicle power demand of the electric vehicle, constructs a function to minimize the energy consumption cost of fuel cell hybrid heavy-duty trucks, including: The voltage degradation of the multi-stack fuel cells is obtained and characterized to construct a multi-stack fuel cell performance degradation model. The battery capacity decay of the power battery is obtained and characterized to construct a power battery performance degradation model. Based on the multi-stack fuel cell performance degradation model and the power battery performance degradation model, the energy consumption cost of fuel cell hybrid heavy-duty truck is determined. The energy consumption cost of fuel cell hybrid heavy-duty truck includes the hydrogen consumption cost of multi-stack fuel cells, the equivalent hydrogen consumption cost of power batteries, the performance degradation cost of multi-stack fuel cells, and the performance degradation cost of power batteries. Based on the total power requirement of the electric vehicle, construct the battery output power constraint conditions; The energy consumption cost of the fuel cell hybrid heavy-duty truck is transformed into a minimization problem function, and the battery output power constraint is introduced to construct the energy consumption cost minimization function of the fuel cell hybrid heavy-duty truck.

4. The method according to claim 3, characterized in that, The specific expression for the function that minimizes the energy consumption cost of the fuel cell hybrid heavy-duty truck is shown below: In the above formula, The function representing the minimization problem of energy consumption cost for fuel cell hybrid heavy-duty trucks is... This indicates the hydrogen consumption cost of multiple fuel cell stacks. This indicates the equivalent hydrogen consumption cost of the power battery. This indicates the cost of performance degradation in multi-stack fuel cells. This indicates the cost of power battery performance degradation. express time, Indicates the total time.

5. The method according to claim 3, characterized in that, The specific expression for the battery output power constraint is as follows: In the above formula, Indicates the whole vehicle is Power required at any time and These represent the minimum and maximum allowable output power of a multi-stack fuel cell, respectively. and These represent the minimum and maximum power output allowed by the power battery, respectively. and This indicates the minimum and maximum allowable SOC values ​​for the power battery. express The output power of multiple fuel cells at any time express The output power of the power battery at all times This indicates the SOC value of the power battery.

6. A tiered energy management system for fuel cell hybrid heavy-duty trucks, characterized in that, The system includes: The first module is used to construct a dynamic model of the electric vehicle and obtain the total power requirement of the electric vehicle. The second module is used to consider the performance degradation of multiple fuel cells and the performance degradation of power batteries, and to construct a function to minimize the energy consumption cost of fuel cell hybrid heavy trucks by combining the overall vehicle power demand of the electric vehicle. The third module is used to transform the energy consumption cost minimization function of the fuel cell hybrid heavy truck and solve it through a deep deterministic strategy gradient algorithm to obtain the power allocation scheme of the multi-stack fuel cell and the power allocation scheme of the power battery. The fourth module is used to perform power distribution through a virtual droop control method based on the power distribution scheme of the multi-stack fuel cells and the power distribution scheme of the power battery, thereby completing the hierarchical energy management of fuel cell hybrid heavy trucks. The process of transforming the energy consumption cost minimization function of the fuel cell hybrid heavy-duty truck and solving it using a deep deterministic strategy gradient algorithm to obtain the power allocation scheme of the multi-stack fuel cells and the power battery includes: By setting state variables, action variables, and a reward function, the function for minimizing the energy consumption cost of the fuel cell hybrid heavy-duty truck is transformed to obtain a Markov game problem for the energy consumption cost of the fuel cell hybrid heavy-duty truck. The state variables include the total power demand of the electric vehicle, the voltage decay of the multiple fuel cells, the capacity decay of the power battery, and the SOC value of the power battery. The action variables include the output power of the multiple fuel cells and the output power of the power battery. The reward function is the energy consumption cost of the fuel cell hybrid heavy-duty truck. The interaction data between the energy consumption cost Markov game problem of the fuel cell hybrid heavy truck and the deep deterministic policy gradient algorithm is obtained, and the deep deterministic policy gradient algorithm is trained through the interaction data to obtain the trained deep deterministic policy gradient algorithm. The current state of the electric vehicle is obtained and input into the trained deep deterministic policy gradient algorithm to obtain the current action of the electric vehicle. The current action of the electric vehicle includes the output power of the multi-stack fuel cell and the output power of the power battery. The process of obtaining the current state of the electric vehicle and inputting it into the trained deep deterministic policy gradient algorithm to obtain the current action of the electric vehicle continues until the electric vehicle stops running. The power allocation scheme of the multi-stack fuel cell and the power allocation scheme of the power battery are then output. The specific expression for the Markov game problem concerning the energy consumption cost of fuel cell hybrid heavy-duty trucks is as follows: In the above formula, Represents state variables. Indicates the amount of action. Represents the reward function, This indicates the total power requirement of an electric vehicle. This indicates the voltage decay of multiple fuel cell stacks. This indicates the amount of battery capacity degradation in a power battery. This indicates the SOC value of the power battery. This indicates the output power of a multi-stacking fuel cell. Indicates the output power of the power battery. This indicates the hydrogen consumption cost of multiple fuel cell stacks. This indicates the equivalent hydrogen consumption cost of the power battery. This indicates the cost of performance degradation in multi-stack fuel cells. This indicates the cost of power battery performance degradation; The step of allocating power according to the power allocation scheme of the multi-stack fuel cells and the power allocation scheme of the power battery, and completing the hierarchical energy management of fuel cell hybrid heavy-duty trucks through a virtual droop control method, includes: The output power of the multi-fuel cell stack is obtained based on the power distribution scheme of the multi-fuel cell stack, and the change in output power of the multi-fuel cell stack is calculated. Based on the output power variation of multiple fuel cells, and combined with the voltage decay of the multiple fuel cells, the output power variation of several fuel cells is determined. Based on the changes in the output power of the aforementioned stacks, determine the reference values ​​for the output power of the stacks, and calculate the virtual droop resistance value. The power distribution of several fuel cells is obtained by adjusting the current of the virtual droop resistor using the virtual droop control method. By combining the power allocation of the aforementioned fuel cell stacks with the power allocation scheme of the power battery, the hierarchical energy management of the fuel cell hybrid heavy-duty truck is completed. The specific expression for calculating the virtual droop resistance value is as follows: In the above formula, Indicates the first The voltage value of the individual fuel cell output voltage after passing through the DC / DC converter. Indicates the first The voltage value at the output terminal of a DC / DC converter after passing through a virtual resistor. Indicates the first The circuit resistance value of each fuel cell stack. Indicates the first The virtual resistor value added to the output of a DC / DC converter. Indicates the first The current at the output of each DC / DC converter, where... , It refers to electric vehicles.

Images