A centerless distributed battery equalization device and control method

Abstract

The application discloses a centerless distributed battery equalization device and a control method, the device is composed of a plurality of battery equalization units, each unit has the ability of battery state detection, charge and discharge control, calculation and communication with other modules. The communication between the units forms a centerless distributed collaborative network, forming a centerless distributed multi-agent system, each unit can use local information to derive a reasonable equalization strategy, and charge and discharge control is performed on the battery monomer according to the equalization strategy, so that all battery monomers are balanced. The application also discloses an improved consensus algorithm, which solves the problem of slow convergence time of the classical consensus algorithm.

Description

Technical Field

[0001] This invention relates to the field of battery balancing technology, specifically to a decentralized distributed battery balancing device and control method. Background Technology

[0002] Current battery balancing devices are limited in design to the number of individual battery cells they can balance simultaneously. A single balancing device can only balance a limited number of cells in a given group at a time. When the number of cells to be balanced is large, a single balancing device cannot complete the task in a single operation. If multiple devices are used for simultaneous balancing or a single device performs multiple operations, the result may be that each batch of cells reaches a different balancing state. This not only multiplies the time required for rebalancing but also increases the complexity of the operation. This inconsistency affects battery pack performance, potentially shortening battery life, increasing maintenance costs, and reducing overall system efficiency. Therefore, developing a solution capable of efficiently balancing a large number of battery cells simultaneously is crucial for maintaining the performance and extending the lifespan of battery packs composed of numerous individual cells.

[0003] In addition, current battery balancing algorithms are computationally complex, requiring the Nash balancing algorithm to combine the power information of all individual cells in the battery pack to derive the balancing strategy. This requires complex calculations from devices such as processors and memory, and the computational efficiency is also reduced.

[0004] Traditional consensus algorithms suffer from a problem where the consensus process slows down over time. As the consensus process progresses, the deviation between control targets gradually decreases. According to the principle of consensus algorithms, the reduction in deviation leads to a gradual slowdown in the convergence speed of the overall consensus of the controlled targets. Therefore, how to adopt a more efficient algorithm to speed up the battery equalization process is also a key issue. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a decentralized distributed battery balancing device and control method. Each battery balancing unit is connected to a single battery cell, thereby realizing a decentralized, distributed battery balancing system based on a consensus algorithm, which solves the problems of slow convergence time and poor scalability of the classic consensus algorithm.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A decentralized distributed battery balancing device,

[0008] It includes multiple battery balancing units. The battery balancing device is decentralized and distributed. Each battery balancing unit is regarded as an intelligent agent node, and the intelligent agent nodes communicate with each other.

[0009] Each of the battery equalization units is connected to one battery cell;

[0010] Each of the battery equalization units includes module M1 and module M2;

[0011] The module M1 includes a battery status detection unit and a communication unit;

[0012] The battery status detection unit is used to detect the status information of the battery cells connected to the battery equalization unit; the communication unit is used to send the status information of the battery cells detected by the battery status detection unit and to receive the status information sent by the adjacent battery equalization unit.

[0013] The module M2 includes a computing unit and a charge / discharge control unit;

[0014] The calculation unit is used to calculate the balancing strategy of the battery cells connected to the battery balancing unit based on the status information detected and received by the module M1; the control unit is used to control the charging and discharging current of the battery cells according to the balancing strategy.

[0015] Preferred,

[0016] Each of the battery balancing device units is equivalent to an intelligent agent node. n Each of the aforementioned battery balancing device units has n There are 1 intelligent agent nodes, and the set of intelligent agent nodes is represented as The edge set in the multi-agent system is represented as: ,in Represents intelligent agent nodes and agent nodes Adjacent to each other and having a communication channel, for agent nodes Its neighborhood set is represented as ;

[0017] The Laplace matrix of the multi-intelligent system , its first OK Column elements are

[0018] ;

[0019] in, Represents a node With nodes Connection weights between them

[0020] ;

[0021] in, These are nodes and nodes The number of adjacent nodes.

[0022] Preferred,

[0023] The state information of the individual battery cell is the battery voltage.

[0024] A battery balancing control method using any of the above-mentioned battery balancing devices, the control method comprising the following steps:

[0025] S1: The battery equalization unit collects the voltage information of the individual cells connected to it at each sampling time;

[0026] S2: Adjacent battery equalization units exchange the collected battery voltage information with each other;

[0027] S3: The battery balancing unit calculates the charging and discharging control strategy of the battery cells connected to it according to the improved consensus algorithm, and then controls the charging and discharging current of the battery cells according to the charging and discharging strategy.

[0028] S4: Repeat S1~S3 until all individual cell voltages reach the same level.

[0029] Preferred,

[0030] The improved consensus algorithm is as follows:

[0031] ;

[0032] In the formula For multi-agent State variables; For its control input; This is the initial time. For convergence time; For the Laplace matrix in (1) above L The smallest non-zero eigenvalue;

[0033] Discretizing it yields:

[0034] ;

[0035] In the formula Represents intelligent agent nodes exist The state variable at any given time, i.e., the agent The state of the detected battery cells Represents intelligent agent nodes exist The state variable at any given time.

[0036] Preferred,

[0037] The intelligent agent node The equilibrium strategy is:

[0038] ;

[0039] In the formula Represents a node exist The change in state at any given time is the balancing strategy for battery cell i.

[0040] Preferred,

[0041] The intelligent node The charging and discharging current of the connected battery cells is controlled according to the equalization strategy as follows:

[0042] ;

[0043] In the formula, According to Determine the proportional factor for the specific charging and discharging current; Represents a node exist The magnitude of the charging and discharging current of the connected individual battery cells at all times.

[0044] Preferred,

[0045] The improved consensus algorithm is executed by each of the intelligent agent nodes independently calculating locally, and then directly controlling the charging and discharging of the battery cells.

[0046] The beneficial effects of this invention are:

[0047] 1. Compared with traditional battery balancing equipment, the battery balancing device and control method of the present invention have the advantages of flexibility and scalability. It can select the number of corresponding battery balancing units according to the number of batteries to be balanced, and then simultaneously perform balancing operations on all batteries to be balanced and balance them to the same state.

[0048] 2. The strategy of this invention is based on a consensus algorithm, ensuring that at any given time, the amount of electricity released by the discharging battery under the balancing strategy is equal to the amount of electricity received by the charging battery. Therefore, this balancing strategy does not consume energy from an external power source during its overall operation.

[0049] 3. The battery balancing method of the improved consensus algorithm of the present invention can combine a decentralized distributed system to achieve battery state balancing through distributed communication and computation. Each step is implemented in a distributed manner, and control is directly performed after each computation step. The convergence time is reduced compared with the classic consensus algorithm.

[0050] 4. The improved consistency control algorithm adopted in this invention can be implemented through distributed communication between battery cells and relatively simple computational iterations, which requires low computing power from the equipment.

[0051] 5. The improved consistency control algorithm adopted in this invention can run in a centralized system in the form of matrix operations, as well as in a distributed system. It also has a rigorously proven convergence property. Based on this, the battery balancing maintenance method is applicable to various maintenance objectives and can flexibly adapt to battery balancing scenarios of different scales.

[0052] 6. This invention also possesses other advantages of a decentralized, distributed system:

[0053] Fault tolerance: Decentralized topology has no single point of failure because information transmission is distributed. Even if some nodes in the network fail, other nodes can still continue to communicate and work, which improves the stability and reliability of the network; due to its fault tolerance, the requirements for communication are low.

[0054] Load balancing: Information processing is distributed across various nodes, so a decentralized topology can achieve balanced computing load, which can avoid overloading of the central node compared to the traditional centralized topology.

[0055] Autonomy: Each node can make decisions and operate independently without the need for instructions from the central node, which allows each part of the network to react quickly based on local information.

[0056] Algorithm Applicability: This invention employs a distributed control method based on an improved consensus algorithm, which has rigorously proven convergence properties. The battery balancing maintenance method based on this algorithm is applicable to various maintenance objectives, such as maintaining battery capacity balancing.

[0057] Scalability: The equalization device of the consistency control algorithm adopted in this invention can adapt to the equalization requirements of battery packs with different numbers of individual cells by adding or removing equalization modules. Attached Figure Description

[0058] The present invention includes the following figures:

[0059] Figure 1 Schematic diagram of the battery equalization unit of this invention

[0060] Figure 2 Schematic diagram of the balanced topology of this invention

[0061] Figure 3 Detailed schematic diagram of the balanced topology of this invention

[0062] Figure 4 Equalized topology schematic diagram of an embodiment of the present invention

[0063] Figure 5Communication topology diagram of a multi-agent system provided in this invention

[0064] Figure 6 Flowchart of the method proposed in this invention

[0065] Figure 7 Convergence speed diagram of traditional consensus algorithm

[0066] Figure 8 The convergence speed diagram of the improved consensus algorithm of this invention. Detailed Implementation

[0067] The present invention will be further described in detail below with reference to the accompanying drawings. The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] This invention provides a decentralized distributed battery balancing device, which consists of multiple battery balancing units, the number of which is equal to the number of batteries to be balanced. Each battery balancing unit is connected to one battery cell. Each battery balancing unit has the following functions: it can connect to one battery cell and collect the voltage state information of the connected battery cell; each battery balancing unit is equivalent to an intelligent agent, and these agents can communicate with each other to form a multi-agent system. Based on the collected and received information, a charging and discharging strategy is calculated according to the algorithm of this invention, and charging and discharging operations are performed on each battery cell according to the strategy.

[0069] Figure 1 This is a schematic diagram of a battery balancing unit in a decentralized distributed battery balancing device of the present invention. The battery balancing unit consists of module M1 (including a battery state detection and communication unit) and module M2 (including a calculation unit and a charge and discharge control unit). Each battery balancing unit is connected to a single battery cell to be balanced and exchanges information with the outside world.

[0070] The specific functions of each module are as follows:

[0071] Module M1: Includes a battery state detection and communication unit, which is responsible for detecting the state information of individual battery cells and communicating with adjacent battery equalization units. The detected and received information serves as the input for subsequent consistency control algorithm calculations.

[0072] Detection unit: Measure the state information (voltage) of individual battery cells according to the set equalization target;

[0073] Communication unit: Communicates with adjacent battery equalization units, sends the status information of individual battery cells detected by the detection unit, and receives status information sent by adjacent battery equalization units.

[0074] Module M2: Includes a calculation unit and a charge / discharge control unit. Based on the state data of the individual battery cells and the state data sent by the adjacent battery equalization units, it calculates the equalization strategy of the individual battery cells connected to this unit through a consistency control strategy, and controls the equalization charging or discharging current of the battery according to the equalization strategy.

[0075] Calculation unit: Calculates the equalization strategy of a battery cell according to the battery equalization maintenance method based on the consistency control algorithm proposed in this invention, based on the detected and received information, that is, the charging / discharging current of the battery cell in the next period.

[0076] Control unit: Controls the charging and discharging current of individual battery cells according to the charging and discharging strategy determined by the computing unit.

[0077] Figure 2 , 3 These are all topology diagrams of a decentralized distributed battery balancing maintenance system composed of the above-mentioned units. The decentralized distributed battery balancing device proposed in this invention can be described as a multi-agent system. The structure of the corresponding multi-agent system is decentralized and distributed. In this system, each battery balancing unit is equivalent to an agent node. The communication between units is regarded as the communication of the multi-agent system and is represented by the edges between nodes.

[0078] This invention considers multiple battery equalization units as a multi-agent system, which has the following characteristics: Each intelligent agent (battery balancing unit), i.e. The set of nodes is represented as _ nodes. The number of nodes is represented as the edge set in a multi-agent system. ,in Represents a node and nodes There is a communication channel, and the two nodes are considered adjacent topologically. For the node... Its neighborhood set is represented as .

[0079] For the entire multi-agent system, define the Laplace matrix. , its first OK Column elements are

[0080] ;

[0081] in, Represents a node With nodes Connection weights between them

[0082] ;

[0083] in, These are nodes and nodes Number of adjacent nodes.

[0084] In this embodiment, the number of nodes is used. For example, the topology diagram is attached. Figure 4 and attached Figure 5 As shown. At this time ;

[0085] ;

[0086] A battery equalization control method based on an improved consensus algorithm, such as Figure 6 As shown, the specific implementation steps are as follows:

[0087] Step 1: The battery equalization unit collects the voltage information of individual cells at each sampling time;

[0088] Based on the set balancing objective, this embodiment selects voltage balancing, reads the current voltage state value of the single battery connected to each balancing unit in the multi-agent system, and uses it as the input of the balancing strategy.

[0089] Step 2: Adjacent equalization units exchange the collected battery voltage information;

[0090] After each sampling, the equalization units communicate with each other to transmit the voltage status data measured by each equalization unit to other adjacent equalization units.

[0091] Step 3: Each battery equalization unit calculates the charge and discharge control strategy of the connected battery cells according to the improved consensus algorithm; then controls the charge and discharge current of the battery cells according to the charge and discharge strategy.

[0092] Each equalization unit combines the voltage state of the individual battery cell it detects with the voltage data received from other equalization units, and uses an improved consensus algorithm to determine the equalization strategy for the connected individual battery cells.

[0093] The improved consensus algorithm is as follows:

[0094] The communication topology of a multi-agent system can be represented by an undirected connected graph. The Laplace matrix is ​​positive semi-definite, and the smallest non-zero eigenvalue of the Laplace matrix is ​​denoted as . Therefore, in this invention, for any node The consensus algorithm for accelerating step size in continuous scenarios is as follows:

[0095] ;

[0096] In the formula For state variables; For control input; and This is the initial time. The set convergence time; For the aforementioned Laplace matrix L The smallest non-zero eigenvalue.

[0097] Discretizing it yields:

[0098] ;

[0099] In the formula Represents a node exist The state quantity at any given time. Represents a node exist The state quantity at any given moment.

[0100] node The equilibrium strategy is:

[0101] ;

[0102] In the formula Represents a node exist The change in state at any given time is the equilibrium strategy.

[0103] node The charging and discharging current of the battery cells is controlled according to the equalization strategy as follows:

[0104] ;

[0105] In the formula, According to Determine the proportional factor for the specific charging and discharging current; Represents a node exist The magnitude of the charging and discharging current at any given moment.

[0106] It is important to note that the algorithm is executed independently by each node locally, and the charging and discharging of individual battery cells is directly controlled after the calculation.

[0107] Step 4: Repeat steps 1 through 3 until all individual cell voltages reach the same level.

[0108] As the equilibrium process continues, until the node... The change amount is zero. In the improved consensus algorithm, the states of all individual cells can be determined within a set time. Internal equilibrium is achieved.

[0109] This embodiment compares the convergence speed of the traditional consensus algorithm and the improved consensus algorithm. Figure 7 To improve the convergence speed of traditional consensus algorithms, see attached... Figure 8 To improve the convergence speed of the consensus algorithm, it can be seen that the convergence speed of the improved consensus algorithm is significantly faster than that of the traditional consensus algorithm.

[0110] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0111] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A battery balancing control method for a battery balancing device, characterized in that, A decentralized distributed battery balancing device includes multiple battery balancing units. The battery balancing device is decentralized and distributed, and each battery balancing unit is regarded as an intelligent agent node. The intelligent agent nodes communicate with each other. Each of the battery equalization units is connected to one battery cell; Each of the battery equalization units includes module M1 and module M2; The module M1 includes a battery status detection unit and a communication unit; The battery status detection unit is used to detect the status information of the battery cells connected to the battery equalization unit; the communication unit is used to send the status information of the battery cells detected by the battery status detection unit and to receive the status information sent by the adjacent battery equalization unit. The module M2 includes a computing unit and a charge / discharge control unit; The calculation unit is used to calculate the balancing strategy of the battery cells connected to the battery balancing unit based on the state information detected and received by the module M1. The control unit is used to control the charging and discharging current of the battery cell according to the equalization strategy; Each of the battery balancing device units is equivalent to an intelligent agent node. n Each of the aforementioned battery balancing device units has n There are 1 intelligent agent nodes, and the set of intelligent agent nodes is represented as ; In a multi-agent system, the edge set is represented as ,in Represents intelligent agent nodes and agent nodes Adjacent to each other and having a communication channel, for agent nodes Its neighborhood set is represented as ; The Laplace matrix of the multi-agent system , its first OK Column elements are ; in, Represents a node With nodes Connection weights between them ; in, These are nodes and nodes The number of adjacent nodes; The control method includes the following steps: S1: The battery equalization unit collects the voltage information of the individual cells connected to it at each sampling time; S2: Adjacent battery equalization units exchange the collected battery voltage information with each other; S3: The battery balancing unit calculates the charging and discharging control strategy of the battery cells connected to it according to the improved consensus algorithm, and then controls the charging and discharging current of the battery cells according to the charging and discharging strategy. S4: Repeat S1~S3 until all individual cell voltages reach the same level; The improved consensus algorithm is as follows: ; In the formula Let be the state variable of multi-agent i; For its control input; This is the initial time. For convergence time; For the Laplace matrix in (1) above L The smallest non-zero eigenvalue; Discretizing it yields: ; In the formula Represents intelligent agent nodes exist The state variable at any given time, i.e., the agent The state of the detected battery cells Represents intelligent agent nodes exist The state variable at any given time; The intelligent agent node The equilibrium strategy is: ; In the formula Represents a node exist The change in state at any given time is the change in state for a single battery cell. The equilibrium strategy.

2. The decentralized distributed battery equalization control method according to claim 1, characterized in that, The state information of the individual battery cell is the battery voltage.

3. The battery equalization control method according to claim 1, characterized in that, The intelligent agent node The charging and discharging current of the connected battery cells is controlled according to the equalization strategy as follows: ; In the formula, According to Determine the proportional factor for the specific charging and discharging current; Represents a node exist The magnitude of the charging and discharging current of the connected individual battery cells at all times.

4. The control method according to claim 1, characterized in that, The improved consensus algorithm is executed by each of the intelligent agent nodes independently calculating locally, and then directly controlling the charging and discharging of the battery cells.

Images