Battery energy optimization management method based on artificial intelligence algorithm

Through the battery energy optimization management method based on artificial intelligence algorithms, the logic gate neural network model is used to solve the problem of battery cell imbalance, and intelligent charging and discharging control is realized, extending the battery system life and improving performance.

CN118888884BActive Publication Date: 2024-11-29SHAANXI WINDRIDERPOWER CO LTD +2
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411397286.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-09
Publication Date
2024-11-29
Estimated Expiration
2044-10-09

AI Technical Summary

Technical Problem

Traditional battery charging and discharging control methods cannot effectively solve the problem of imbalance between single batteries, resulting in increased battery aging and safety risks.

Method used

Adopt battery energy optimization management method based on artificial intelligence algorithms, and build a neural network model based on logic gates, collect battery state variable parameters, execute logical formula judgments, and generate the optimal charging and discharge decision strategy to ensure that the battery cell operates within a safe range. and battery balance.

Benefits of technology

It realizes intelligent charging and discharging control strategy optimization, improves control accuracy and effectiveness, extends the battery system life, improves battery pack performance, and has good scalability and flexibility.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118888884B_ABST
    Figure CN118888884B_ABST
Patent Text Reader

Abstract

The embodiment of the present application provides a battery energy optimization management method based on an artificial intelligence algorithm, which relates to the field of artificial intelligence technology. The method includes: constructing a battery energy optimization management model based on a neural network of logic gates; collecting state variable parameters of each battery cell in the battery system; inputting the state variable parameters of each battery cell into the battery energy optimization management model, and the battery energy optimization management model performs judgment based on a logic formula, and makes charging and discharging decisions through logical operations combined with neuron weights to obtain the optimal charging and discharging decision strategy for each battery cell; executes the optimal charging and discharging decision strategy for each battery cell, and finally outputs the charging current and discharging current of each battery cell to ensure that the battery cell operates within a safe range and the battery power is balanced. Through this solution, the intelligent charging and discharging control strategy optimization of the battery system can be realized, the battery system life can be extended, and the performance of the battery pack can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a battery energy optimization management method based on artificial intelligence algorithm. Background Art

[0002] With the rapid development of battery technology, lithium-ion batteries, lead-acid batteries, etc. have been widely used in various electronic devices, electric vehicles and energy storage systems. Battery systems are usually composed of multiple battery cells. The charge and discharge states of individual batteries are often different. Long-term unbalanced charge and discharge may lead to increased battery aging and even safety problems. In order to improve the service life and efficiency of battery systems, balanced charge and discharge strategies have become an important research direction in battery management systems. Traditional battery charge and discharge control methods usually rely on simple current control or algorithms based on empirical rules, but these methods often cannot effectively deal with the imbalance problem of single cells in complex battery systems.

[0003] Therefore, developing an intelligent balanced charging and discharging strategy is of great significance to extending the life of the battery system and improving the performance of the battery pack. Summary of the invention

[0004] In order to address the deficiencies of the prior art, an embodiment of the present application provides a battery energy optimization management method based on an artificial intelligence algorithm to solve the problem that traditional battery charging and discharging control methods generally rely on simple current control or algorithms based on empirical rules and cannot effectively solve the technical problem of imbalance between single cells.

[0005] An embodiment of the present application provides a battery energy optimization management method based on an artificial intelligence algorithm, including: constructing a battery energy optimization management model based on a neural network of logic gates; collecting state variable parameters of each battery cell in a battery system, wherein the state variable parameters include: voltage and remaining power; inputting the state variable parameters of each battery cell into the battery energy optimization management model, wherein the battery energy optimization management model performs judgment based on a logic formula, performs charging and discharging decisions through logical operations combined with neuron weights, and obtains an optimal charging and discharging decision strategy for each battery cell; executes the optimal charging and discharging decision strategy for each battery cell, and finally outputs the charging current and discharging current of each battery cell to ensure that the battery cell operates within a safe range and the battery power is balanced.

[0006] In a possible implementation, the battery energy optimization management model based on a neural network of logic gates is constructed, including: constructing parameter variables of the input neural network, namely, state variables and decision variables, wherein the state variables are the voltage and remaining power of each battery cell, and the decision variables are the charge and discharge decisions of each battery; constructing a backbone network based on logic neurons, wherein the neurons of the backbone network are logic neurons, namely, each neuron represents a logical relationship in the battery system, and the neurons express a bidirectional relationship with each neighbor, allowing reasoning in any direction; constructing a loss function of the backbone network, and training the neural network weights through the loss function; constructing a classification head of the network, which is used to classify and output charge and discharge decisions based on the extracted features.

[0007] In one possible implementation, the state variable includes the voltage of each battery cell. , remaining power ; The decision variables include the charge and discharge decision of each battery, that is, the charging current of each battery cell and discharge current , specifically, Indicates The charging current of each battery cell, Indicates The discharge current of a battery cell.

[0008] In a possible implementation, the construction of a backbone network based on logic neurons includes: constructing three types of logic neurons, namely, battery charge balance neurons, charge and discharge constraint neurons, and voltage balance target neurons; constructing three types of input neurons, namely, first input neurons , the second input neuron and , the third input neuron ; Among them, the first input neuron , used to obtain the state variable information related to the remaining battery power, indicating that the input The current remaining capacity of each battery cell ; The second input neuron includes and , used to obtain the The current charging current of each battery cell and discharge current ; The third input neuron , which is the Voltage of each battery cell The neurons mapped to the input layer are used to represent the input Voltage of each battery cell , and pass it as input to the logic neurons in the middle layer.

[0009] In a possible implementation, the construction of a backbone network based on logic neurons also includes: constructing logic formulas corresponding to various types of logic neurons, thereby establishing a one-to-one correspondence between the logic neurons and the elements of the logic formula, that is, each neuron represents an operation in the logic formula, and the connection and activation functions of the neurons correspond to the connectives and operations in the logic formula.

[0010] In a possible implementation, the constructing of the logic formula includes: constructing a battery power balance formula corresponding to the battery power balance neuron; the battery power balance formula is:

[0011] ,in, Indicates The change of the charge of a battery cell over time is determined by the charging current and the discharging current; Indicates time; the battery power balance neuron is used to execute the power balance formula to calculate the The rate of change of charge of each battery cell And make logical judgments and decide whether the charge and discharge current needs to be adjusted.

[0012] In a possible implementation, the constructing of the logic formula further includes: constructing a charge-discharge constraint formula corresponding to the charge-discharge constraint neuron; the charge-discharge constraint formula is: ,in, Indicates the maximum charging current limit; Indicates the maximum discharge current limit; the charge and discharge constraint neuron is used to indicate the constraints of the charge current and discharge current. Its input is the charge or discharge current of the battery, and its activation function is designed to determine whether the current exceeds the constraint value based on the input information. Specifically, the charge and discharge constraint judgment is performed based on the comparison between the input current value and the maximum current limit value: if the charge current Exceeding the maximum value set , the charge-discharge constraint neuron will activate the corresponding adjustment mechanism to reduce the charging current; if the discharge current Exceeding the maximum value set , the charge and discharge constraint neurons will activate the corresponding adjustment mechanism to reduce the discharge current.

[0013] In a possible implementation, the constructing of the logic formula further comprises: constructing a voltage balance target formula corresponding to the voltage balance target neuron; the voltage balance target formula is:

[0014] ,in, Indicates The voltage of each battery cell, Indicates The voltage of each battery cell; the voltage balance target formula aims to minimize the voltage difference between the battery cells, thereby achieving voltage balance and ensuring that each battery operates within the ideal voltage range; the voltage balance target neuron is used to execute the voltage balance logic and determine the voltage difference between the two battery cells. Is it too big: If Exceeding the set threshold , the neuron will be activated and trigger the corresponding charging and discharging strategy to reduce the voltage difference.

[0015] In one possible implementation, the loss function of the backbone network is constructed, and the neural network weights are trained by the loss function, including: the loss function According to the battery voltage difference and power balance calculate:

[0016] ,in, To minimize the voltage difference between cells, cell balancing This is to ensure that the change in charge matches the current. is a trade-off parameter.

[0017] In a possible implementation, the classification head of the constructed network is used to classify and output charging and discharging decisions according to the extracted features, including: outputting charging and discharging decisions, by adjusting the weights and biases of all logical neurons so that the network fully connected layer generates the optimal charging and discharging strategy according to the input battery status information, and the final output of the network is the charging current and discharging current of each battery cell:

[0018] ,in, is the output vector of the network, which is expressed as:

[0019] , represents the optimal strategy of charging and discharging at the current moment; is the number of battery cells; Indicates the output The charging current of each battery cell; Indicates the output The discharge current of each battery cell; is the input variable, which contains the battery status information. The input vector is in the form of:

[0020] ; is the weight of each input variable, indicating the impact of each battery cell on the decision; is the bias term, which is used to adjust the output value of the neuron; is the ReLU activation function.

[0021] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the battery energy optimization management method based on artificial intelligence algorithm proposed in the present invention can not only learn the complex nonlinear relationship in the battery system, but also express the charging and discharging constraints and balancing requirements between battery cells through logical formulas, thereby realizing intelligent optimization of charging and discharging control strategies, ensuring that the battery cells operate within a safe range and the battery power is balanced. The artificial intelligence algorithm of the present invention enables the charging and discharging control strategy to be adaptively adjusted according to the actual state of the battery, improves the accuracy and effectiveness of the control, and effectively solves the imbalance problem between single cells, thereby extending the life of the battery system and improving the performance of the battery pack. In addition, the present invention is based on neural network construction and has good scalability and flexibility. When the scale or complexity of the battery system increases, the structure and parameters of the neural network can be adjusted to adapt to new requirements. At the same time, this method can also be combined with other battery management technologies to form a more complete battery management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A flowchart of a battery energy optimization management method based on an artificial intelligence algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0025] Those skilled in the art will appreciate that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, etc., and do not represent any specific technical meaning, nor do they represent the necessary logical order between them. It should also be understood that in the embodiments of the present application, "multiple" may refer to two or more, and "at least one" may refer to one, two or more. It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, in the absence of explicit limitation or contrary revelation given in the context, it can generally be understood as one or more. In addition, the term "and / or" in the present application is only a kind of association relationship describing the associated objects, indicating that there can be three kinds of relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present application generally indicates that the associated objects before and after are an "or" relationship. It should also be understood that the description of each embodiment in the present application emphasizes the differences between the embodiments, and the same or similar parts can refer to each other. For the sake of brevity, they will not be repeated one by one.

[0026] At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered as part of the specification. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0027] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0028] The present invention proposes a battery energy optimization management method based on an artificial intelligence algorithm. The battery energy optimization management model constructed based on a neural network of logic gates can not only learn the complex nonlinear relationships in the battery system, but also express the charging and discharging constraints and balancing requirements between battery cells through logical formulas, thereby realizing intelligent optimization of the charging and discharging control strategy.

[0029] Figure 1The present invention provides a flowchart of a battery energy optimization management method based on an artificial intelligence algorithm. Figure 1 As shown, a battery energy optimization management method 100 based on artificial intelligence algorithm includes the following steps: S110: constructing a battery energy optimization management model based on a neural network of logic gates.

[0030] Specifically, in step S110, a battery energy optimization management model based on a neural network of logic gates is constructed, including the following steps: S111: constructing parameter variables of the input neural network, namely, state variables and decision variables, wherein the state variables are the voltage and remaining power of each battery cell, and the decision variables are the charge and discharge decisions of each battery.

[0031] Specifically, the state variables include the voltage of each battery cell , remaining power ; The decision variables include the charge and discharge decision of each battery, that is, the charging current of each battery cell and discharge current , specifically, Indicates The charging current of each battery cell, Indicates The discharge current of a battery cell.

[0032] S112: Construct a backbone network based on logic neurons, wherein the neurons of the backbone network are logic neurons, that is, each neuron represents a logical relationship in the battery system, and the neuron expresses a bidirectional relationship with each neighbor, allowing reasoning in any direction; the neural network of the present invention is different from the traditional neural network.

[0033] Among them, a backbone network based on logic neurons is constructed to construct three types of logic neurons, namely battery charge balance neurons, charge and discharge constraint neurons, and voltage balance target neurons.

[0034] Construct three types of input neurons, namely the first input neuron , the second input neuron and , the third input neuron .

[0035] The first input neuron , used to obtain the state variable information related to the remaining battery power, indicating the The current remaining capacity of each battery cell ; The second input neuron includes and , used to obtain the The current charging current of each battery cell and discharge current ; The third input neuron , which is the Voltage of each battery cell The neurons mapped to the input layer are used to represent the Voltage of each battery cell , and pass it as input to the logic neurons in the middle layer.

[0036] Optionally, in some embodiments, constructing a backbone network based on logic neurons further includes: constructing logic formulas corresponding to various types of logic neurons, thereby establishing a one-to-one correspondence between the logic neurons and the elements of the logic formula, that is, each neuron represents an operation in the logic formula, and the connection and activation functions of the neurons correspond to the connectives and operations in the logic formula.

[0037] In this embodiment, the activation function of the neuron is defined by a logical formula. The weights that originally played a specific role in the traditional neural network no longer exist in the logical neural network, and are replaced by logic gates. These logic gates are responsible for connecting and processing information in the network, and the logical rules are expressed through symbolic logic. Judgments based on logical formulas are executed, and charging and discharging decisions are made through logical operations combined with neuron weights.

[0038] Specifically, constructing a logic formula of a backbone network based on logic neurons includes the following parts: (1) constructing a logic formula includes: constructing a battery power balance formula corresponding to a battery power balance neuron.

[0039] The battery balance formula represents the change of battery charge over time, which is determined by the charging current and the discharging current. The change of battery charge is equal to the charging current minus the discharging current. The values ​​of the charging current and the discharging current can be passed as input signals to the battery balance neuron. The activation function of the battery balance neuron will calculate the change of charge based on these input signals and the weight of the battery balance neuron. This calculation process can simulate the relationship of the power balance formula, so that the behavior of the neuron is consistent with the physical law of power balance.

[0040] Specifically, the battery charge balance formula is:

[0041] ,in, Indicates The change of the charge of a battery cell over time is determined by the charging current and the discharging current; Indicates The power of each battery cell (unit: Ah); Indicates The charging current of each battery cell (unit: A); Indicates The discharge current of each battery cell (unit: A); represents time (unit: hour). Among them, the first input neuron , which is responsible for obtaining the state variable information related to the remaining battery power, indicating the The current remaining capacity of each battery cell ; The second input neuron includes and , used to obtain the The current charging current of each battery cell and discharge current ; A battery power balance neuron, used to execute the power balance formula to calculate the first The rate of change of charge of each battery cell And make logical judgments and decide whether the charge and discharge current needs to be adjusted. The calculation formula implemented by neurons is expressed as:

[0042] , represents a battery balance neuron. If the neuron If the output is too large or too small, it means that the battery charge change rate is not as expected and needs to be adjusted.

[0043] Charge and discharge current adjustment: If the charge change of a battery is not within the expected range, the logic neuron will trigger the output neuron to adjust the current. If the battery charge is too low, the logic neuron will trigger the charging current to increase. The formula is:

[0044] On the contrary, when the charge is too high, it will trigger an increase in the discharge current.

[0045] Output neuron: The output neuron controls the charge and discharge current of each battery according to the judgment of the logic neuron, i.e., the battery power balance neuron. Output neuron and The charging or discharging current will be adjusted according to the demand for power balance to ensure the balance of battery power.

[0046] Activation function: Logical neurons, i.e., battery power balance neurons, realize power balance control through activation functions, and judge according to the threshold value of power change rate:

[0047] , if the battery charge change rate deviates from the target value, the charge and discharge adjustment operation is activated. (2) Constructing the logic formula also includes: constructing the charge and discharge constraint formula corresponding to the charge and discharge constraint neuron; the charge and discharge constraint formula is:

[0048] ,in, Indicates the maximum charging current limit (unit: A); Indicates the maximum discharge current limit (unit: A).

[0049] The charge and discharge constraint formula indicates that in order to protect the battery, the charge and discharge current cannot exceed the maximum value allowed by the device, and the upper limit of the charge and discharge current is defined to ensure the safe operation of the battery and the device. The constraints of the charge and discharge current are represented by the charge and discharge constraint neurons, and the input of these charge and discharge constraint neurons is the charge or discharge current of the battery. The activation function of the charge and discharge constraint neuron is designed to determine whether the current exceeds the constraint value based on the input information.

[0050] Second input neuron: responsible for obtaining the current charging and discharging current of each battery cell. Second input neuron and Read the actual charge and discharge current of the battery.

[0051] Charge and discharge constraint neuron: used to represent the constraints of charging current and discharging current. Its input is the charging or discharging current of the battery, and its activation function is designed to judge whether the current exceeds the constraint value based on the input information. Specifically, the charge and discharge constraint judgment is performed based on the comparison between the input current value and the maximum current limit value.

[0052] For example, logic neurons Will determine whether the charging current exceeds the maximum value:

[0053] , if the charging current Exceeding the maximum value set , the charge and discharge constraint neurons will activate the corresponding adjustment mechanism to reduce the charging current.

[0054] Similarly, the discharge current will also have similar logical judgment. , logic neurons will also perform similar logic:

[0055] , if the discharge current Exceeding the maximum value set , the charge and discharge constraint neurons will activate the corresponding adjustment mechanism to reduce the discharge current.

[0056] These logical judgments determine whether charging and discharging are operating within a safe range through activation functions and neuron outputs.

[0057] Output neurons: Output neurons generate adjustment signals based on logical judgment results to ensure that the current is within a safe range. Exceeding the maximum value, output neuron Will reduce the charging current:

[0058] , for the discharge current, if Exceeding the maximum value, output neuron Will reduce the discharge current.

[0059] The activation function is used to determine whether to perform charge and discharge adjustment. For charge and discharge constraints, the activation function can be set to be activated based on the comparison between the charge and discharge current and the maximum limit value.

[0060] Activation function: When the current exceeds the maximum limit, the activation signal is 0, indicating that the current needs to be reduced; otherwise, the activation signal is 1, indicating normal operation:

[0061] ,Similarly, the activation function of discharge is also determined by comparing the current with the maximum discharge value.

[0062] (3) Constructing the logic formula also includes: constructing a voltage balance target formula corresponding to the voltage balance target neuron; wherein the voltage balance target formula is:

[0063] ,in, Indicates The voltage of each battery cell (unit: V), Indicates The voltage of a battery cell (unit: V).

[0064] The goal of this voltage balance target formula is to minimize the voltage difference between battery cells, thereby achieving voltage balance and ensuring that each battery operates within the ideal voltage range.

[0065] The third input neuron , which is the Voltage of each battery cell The neurons mapped to the input layer are used to represent the Voltage of each battery cell , and pass it as input to the logic neurons in the middle layer.

[0066] Voltage balancing target neurons, used to execute voltage balancing logic and determine the voltage difference between two battery cells Is it too large?

[0067] A logic neuron is a voltage-balanced target neuron It can calculate:

[0068] ,if Exceeding the set threshold , the neuron will be activated and trigger the corresponding charging and discharging strategy to reduce the voltage difference.

[0069] Output neurons: Output neurons are responsible for generating control signals for charging and discharging currents. For example, when a logic neuron determines and When the difference between Will increase The battery charging current is increased to reduce the voltage difference.

[0070] Linear activation function: If the voltage difference When a certain threshold is exceeded, the charge and discharge control signal is output:

[0071] , the logical operation is defined as:

[0072] .

[0073] S113: Construct a loss function of the backbone network and train the neural network weights through the loss function.

[0074] In step S113, the neural network weights are trained through the loss function. The loss function provides an optimization target for the backbone network. By measuring the difference between the predicted category and the real category, the parameters of the backbone network are updated along the direction where the loss function decreases fastest, guiding the backbone network to learn feature representations that can better complete specific tasks. By constructing the loss function, the voltage difference and power balance between batteries can be clearly optimized, so that the charging and discharging strategy can more accurately meet the needs of battery management. The introduction of the loss function provides a clear optimization target for the training of the neural network, improving the training efficiency and effect.

[0075] Among them, the loss function According to the battery voltage difference and power balance calculate:

[0076] ,in, To minimize the voltage difference between cells, cell balancing This is to ensure that the change in charge matches the current. is a trade-off parameter.

[0077] Under ideal conditions, the power is balanced .

[0078] S114: Construct a classification head of the network to perform classification and output charging and discharging decisions based on the extracted features.

[0079] In step S114, a classification head of the network is constructed (also called a fully connected layer, whose function is to perform classification based on the extracted features).

[0080] Output charging and discharging decisions, by adjusting the weights and biases of all logic neurons, so that the network fully connected layer generates the optimal charging and discharging strategy based on the input battery status information. The final output is the charging current and discharging current of each battery cell. , where y is the output of the network, representing the optimal strategy for charging and discharging at the current moment. The output vector is in the form of:

[0081] .in, is the number of battery cells. Output charging current :Indicates the The charging current of each battery cell. Output discharge current :Indicates the The discharge current of a battery cell.

[0082] is the input variable, which contains the battery status information. The input vector is in the form of:

[0083] Among them, the battery voltage : Indicates the voltage status of each battery cell. Battery power : Indicates the current charge of each battery cell. Current charging current and discharge current : The charge and discharge current values ​​calculated in the previous layers. is the weight of each input, indicating the influence of each battery cell on the decision. is the bias term, which adjusts the output value of the neuron. is the ReLU activation function.

[0084] S120: Collecting state variable parameters of each battery cell in the battery system, where the state variable parameters include: voltage and remaining power.

[0085] S130: Input the state variable parameters of each battery cell into the battery energy optimization management model. The battery energy optimization management model performs judgment based on logic formulas, makes charging and discharging decisions through logical operations combined with neuron weights, and obtains the optimal charging and discharging decision strategy for each battery cell.

[0086] S140: Execute the optimal charge and discharge decision strategy for each battery cell, and finally output the charging current and discharging current of each battery cell to ensure that the battery cell operates within a safe range and the battery power is balanced.

[0087] According to the above-mentioned embodiment of the present invention, the charging and discharging constraints and balancing requirements between battery cells are expressed through logical formulas, thereby realizing intelligent optimization of the charging and discharging control strategy, ensuring that the battery cells operate within a safe range and that the battery power is balanced. This method enables the charging and discharging control strategy to be adaptively adjusted according to the actual state of the battery, thereby improving the accuracy and effectiveness of the control. The constructed logical neurons and logical formulas comprehensively consider multiple aspects such as battery power balance, charging and discharging constraints, and voltage balance, ensuring the safety and stability of the battery during the charging and discharging process. It effectively solves the imbalance problem between single cells, thereby extending the life of the battery system and improving the performance of the battery pack.

[0088] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0089] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0090] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0091] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A battery energy optimization management method based on artificial intelligence algorithm, characterized in that: include: Construct a battery energy optimization management model based on a neural network of logic gates; wherein, the construction of the battery energy optimization management model based on a neural network of logic gates includes: constructing parameter variables of the input neural network, namely, state variables and decision variables, wherein the state variables are the voltage and remaining power of each battery cell, and the decision variables are the charge and discharge decisions of each battery; construct a backbone network based on logic neurons, wherein the neurons of the backbone network are logic neurons, namely, each neuron represents a logical relationship in the battery system, and the neuron expresses a bidirectional relationship with each neighbor, allowing reasoning in any direction; wherein, the construction of the backbone network based on logic neurons includes: constructing three types of logic neurons, namely, battery power balance neurons, charge and discharge constraint neurons, and voltage balance target neurons; constructing three types of input neurons, namely, first input neurons , the second input neuron and , the third input neuron ; Among them, the first input neuron , used to obtain the state variable information related to the remaining battery power, indicating that the input The current remaining capacity of each battery cell ; The second input neuron includes and , used to obtain the The current charging current of each battery cell and discharge current ; The third input neuron , which is the Voltage of each battery cell The neurons mapped to the input layer are used to represent the input Voltage of each battery cell , and pass it as input to the logical neurons in the middle layer; construct the loss function of the backbone network, and train the neural network weights through the loss function; wherein, the loss function of the backbone network is constructed, and the weights of the neural network are trained through the loss function, including: loss function According to the battery voltage difference and power balance calculate: ,in, To minimize the voltage difference between cells, cell balancing This is to ensure that the change in charge matches the current. For trade-off parameters; Indicates The change of the charge of a battery cell over time is determined by the charging current and the discharging current; Indicates time; constructs a classification head of the network, which is used to classify and output charge and discharge decisions according to the extracted features; wherein, the classification head of the network is used to classify and output charge and discharge decisions according to the extracted features, including: outputting charge and discharge decisions, by adjusting the weights and biases of all logical neurons, so that the network fully connected layer generates the optimal charge and discharge strategy according to the input battery status information, and the final output of the network is the charge current and discharge current of each battery cell: ,in, is the output vector of the network, which is expressed as: , represents the optimal strategy of charging and discharging at the current moment; is the number of battery cells; is the input variable, which contains the battery status information. The input vector is in the form of: ; Indicates the output The charging current of each battery cell; Indicates the output The discharge current of each battery cell; is the weight of each input variable, indicating the impact of each battery cell on the decision; is the bias term, which is used to adjust the output value of the neuron; is a ReLU activation function; collecting state variable parameters of each battery cell in the battery system, the state variable parameters including: voltage and remaining power; inputting the state variable parameters of each battery cell into the battery energy optimization management model, the battery energy optimization management model performs judgment based on logic formulas, performs charge and discharge decisions through logic operations combined with neuron weights, and obtains the optimal charge and discharge decision strategy for each battery cell; executes the optimal charge and discharge decision strategy for each battery cell, and finally outputs the charging current and discharging current of each battery cell to ensure that the battery cell operates within a safe range and the battery power is balanced.

2. The method according to claim 1, characterized in that: The state variable is the voltage of each battery cell , remaining power ; The decision variable is the charge and discharge decision of each battery, that is, the charging current of each battery cell. and discharge current , specifically, Indicates The charging current of each battery cell, Indicates The discharge current of a battery cell.

3. The method according to claim 1, characterized in that: in, The construction of the backbone network based on logic neurons also includes: constructing logic formulas corresponding to various types of logic neurons, thereby establishing a one-to-one correspondence between the logic neurons and the elements of the logic formula, that is, each neuron represents an operation in the logic formula, and the connection and activation function of the neurons correspond to the connectives and operations in the logic formula.

4. The method according to claim 3, characterized in that: in, Constructing a logic formula includes: constructing a battery power balance formula corresponding to the battery power balance neuron; the battery power balance formula is: , the battery power balance neuron is used to execute the power balance formula to calculate the first The rate of change of charge of each battery cell And make logical judgments and decide whether the charge and discharge current needs to be adjusted.

5. The method according to claim 4, characterized in that: in, Constructing the logic formula also includes: constructing a charge and discharge constraint formula corresponding to the charge and discharge constraint neuron; the charge and discharge constraint formula is: ,in, Indicates the maximum charging current limit; Indicates the maximum discharge current limit; the charge and discharge constraint neuron is used to indicate the constraints of the charge current and discharge current. Its input is the charge or discharge current of the battery, and its activation function is designed to determine whether the current exceeds the constraint value based on the input information. Specifically, the charge and discharge constraint judgment is performed based on the comparison between the input current value and the maximum current limit value: if the charge current Exceeding the maximum value set , the charge-discharge constraint neuron will activate the corresponding adjustment mechanism to reduce the charging current; if the discharge current Exceeding the maximum value set , the charge and discharge constraint neurons will activate the corresponding adjustment mechanism to reduce the discharge current.

6. The method according to claim 5, characterized in that: in, Constructing a logic formula also includes: constructing a voltage balance target formula corresponding to the voltage balance target neuron; the voltage balance target formula is: ,in, Indicates The voltage of each battery cell, Indicates The voltage of each battery cell; the voltage balance target formula aims to minimize the voltage difference between the battery cells, thereby achieving voltage balance and ensuring that each battery operates within the ideal voltage range; the voltage balance target neuron is used to execute the voltage balance logic and determine the voltage difference between the two battery cells. Is it too big: If Exceeding the set threshold , the neuron will be activated and trigger the corresponding charging and discharging strategy to reduce the voltage difference.

Citation Information

Patent Citations

  • Method for predicating life of aerospace Ni-Cd storage battery based on improved dynamic wavelet neural network

    CN103018673A

  • Optimization operation method for optical diesel store micro-network system with electric vehicle access

    CN107196586A