Special lithium ion battery system for UPS (Uninterrupted Power Supply) based on intelligent management

Abstract

The invention provides a special lithium ion battery system for a UPS (Uninterrupted Power Supply) based on intelligent management, and relates to the technical field of UPSs, the system comprises a battery state monitoring and predicting module, an intelligent charging management module and a fault diagnosis and fault-tolerant control module of a battery system, and the SOH and SOC of a battery are accurately predicted through a support vector machine, a random forest and a particle filter algorithm. A basis is provided for reasonable charging and discharging, the service life of the battery is prolonged, the intelligent charging management module dynamically adjusts charging parameters according to the SOC and the temperature of the battery through a fuzzy logic control algorithm, balance of rapid charging and battery protection is achieved, the intelligent charging management module adapts to different environments, and the fault diagnosis and fault-tolerant control module detects faults in time with the help of a support vector machine algorithm. A model prediction control algorithm optimizes control input when a fault occurs, it is ensured that the UPS system still operates normally under the battery fault, a data acquisition and transmission module and a system monitoring and display module monitor the battery state and provide historical data query analysis, and a user can conveniently master the battery condition and make a decision.

Description

Technical Field

[0001] The invention relates to the technical field of UPS power supply, in particular to a UPS special lithium ion battery system based on intelligent management. Background Art

[0002] According to a UPS battery management system disclosed in China with publication number "CN118238671A", it includes a micro control unit and a plurality of functional modules connected to the micro control unit, the plurality of functional modules include a first group of functional modules and a second group of functional modules, the first group of functional modules are used for battery inspection, and the micro control unit is configured to control the states of the functional modules so that the UPS battery management system has the following working modes: a normal working mode, wherein the micro control unit is configured to turn on the plurality of functional modules; a sleep mode, wherein the micro control unit is configured to turn off the plurality of functional modules; and a timed inspection mode, wherein the micro control unit is configured to turn on only the first group of functional modules on a regular basis on the basis of the sleep mode to inspect the battery. The battery management system of the present invention has perfect functions, can continuously monitor the battery status, has high safety performance, low energy consumption, and is easy to operate.

[0003] The above patent documents and prior art have the following technical problems when used:

[0004] Problem 1: Traditional UPS battery systems usually use simple fuel gauges or experience-based estimation methods to determine the battery's state of charge (SOC) and state of health (SOH). These methods are often not accurate enough, making it difficult for users to accurately understand battery performance and remaining power, affecting battery use and maintenance planning;

[0005] Problem 2: Existing UPS battery charging management usually adopts fixed charging modes, such as constant current charging or constant voltage charging, and cannot be dynamically adjusted according to the actual state of the battery and environmental conditions, resulting in low charging efficiency and shortened battery life. In high or low temperature environments, the performance of the battery will change, but the existing charging system cannot automatically adjust the charging parameters according to temperature changes, which may result in poor charging results or even damage the battery.

[0006] Question three: In the above documents and prior art, the existing UPS battery system has a lag in fault detection, and the fault is often discovered only after it has occurred and caused a certain impact on the system. When a battery fails, the existing system often lacks effective fault-tolerant control measures and cannot adjust the system operating parameters in time to reduce the impact of the fault, which may cause the entire UPS system to fail to work properly. Summary of the invention

[0007] Technical issues solved

[0008] In view of the deficiencies of the prior art, the present invention provides a UPS-specific lithium-ion battery system based on intelligent management, which solves the following problems:

[0009] 1. The battery status prediction is not accurate enough and the UPS battery cannot be adjusted according to the charging environment;

[0010] 2. The UPS battery system has a single charging strategy and poor adaptability to different temperature environments;

[0011] 3. The UPS battery system has untimely fault detection and limited fault tolerance, which leads to the system failure to be handled in time and unable to work.

[0012] Technical Solution

[0013] To achieve the above objectives, the present invention is implemented through the following technical solutions: a UPS-specific lithium-ion battery system based on intelligent management, the system includes a battery status monitoring and prediction module, an intelligent charging management module and a battery system fault diagnosis and fault-tolerant control module, wherein:

[0014] The battery state monitoring and prediction module: firstly, a battery state prediction model is established by using a support vector machine algorithm and a random forest algorithm, and the SOH and SOC of the battery are predicted according to the input real-time data, including voltage, current, and temperature, and the prediction results are used as inputs of a particle filter algorithm, which corrects and optimizes the prediction results by continuously updating the positions and weights of particles;

[0015] The intelligent charging management module: the fuzzy logic control algorithm dynamically adjusts the charging parameters according to the current state SOC and temperature T of the battery. In the early stage of charging, when the SOC is low and the battery temperature is within a suitable range, the fuzzy logic controller outputs a larger charging current value and adopts a constant current charging mode to quickly increase the battery power. As the charging progresses, when the battery is close to being fully charged and the SOC is high or the temperature is too high, the fuzzy logic controller will reduce the charging current and switch to a constant voltage charging mode to ensure that the battery is fully charged and not overcharged, while protecting the battery life;

[0016] The fault diagnosis and fault-tolerant control module of the battery system uses a support vector machine algorithm to analyze the operating data of the battery system to determine whether there is a fault. When a fault is detected, the model predictive control algorithm predicts the future performance based on the mathematical model of the battery system, and optimizes the control input, such as adjusting the charging current and discharging current, to reduce the impact of the fault on the entire system, ensuring that the UPS system can still operate normally in the event of a battery failure. At the same time, the MPC algorithm dynamically adjusts the control strategy according to the development of the fault.

[0017] Preferably, the support vector machine algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps:

[0018] Sp1: Collect a large amount of operating data of UPS-specific lithium-ion batteries under different usage conditions, including voltage , Current ,temperature , Charge and discharge cycle times , and mark these data as the training set:

[0019]

[0020] in, represents the input feature vector, Indicates the corresponding battery status label, including or value;

[0021] Sp2: Determine the optimal hyperplane under constraints by solving the quadratic programming problem , the solution formula is as follows:

[0022]

[0023] The constraints are as follows:

[0024]

[0025] in, is the weight vector, is the bias term, is the penalty parameter used to balance the penalty for interval maximization and misclassification, is the slack variable, Representation vector The norm of , is the number of training samples;

[0026] Sp3: For new input data ,pass Calculate the predicted battery state, represents a symbolic function, if ,but ,like ,but .

[0027] Preferably, the random forest algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps:

[0028] Sp1: Randomly extract multiple sample subsets from the original training data, and the size of each subset is the same as the original training set;

[0029] Sp2: For each sample subset, a decision tree is constructed. When constructing the decision tree, some features are randomly selected as split features at each node, and the optimal split point is selected for splitting;

[0030] Sp3: For new input data, the prediction results of multiple decision trees are integrated and the final prediction result is determined by majority voting, that is:

[0031]

[0032] in, For the final prediction result, Represents input data, is the number of decision trees, For the A decision tree, Indicates majority.

[0033] Preferably, the particle filter algorithm in the battery status monitoring and prediction module corrects and optimizes the prediction result through the following steps:

[0034] Sp1: Initialization: At the initial moment ,generate Particles , and according to the prior distribution Determine the initial weight of each particle ;

[0035] Sp2: Prediction: Based on the system state transition model , predict the position of each particle at the current moment ,in, is the control input;

[0036] Sp3: Update: Based on the observation model and observations , calculate the observation likelihood of each particle , then update the particle weights ;

[0037] Sp4: Resampling: If the effective number of samples of a particle Less than the set threshold , then resample to obtain a new particle set , and reset the weights to ,in, represents the number of particles, Indicates at time No. The weight of each particle;

[0038] Sp5: State estimation: According to the weight and position of the particle, by calculating the moment An estimate of the system state To estimate the system state:

[0039]

[0040] The particle weight update formula is:

[0041]

[0042] in,

[0043] , They are Moment The weight and position of each particle;

[0044] , They are Moment and Moment The weight of each particle;

[0045] At a given particle position In the case of The conditional probability of

[0046] According to the system state transition model, from time The particle position Transfer to the particle position at time probability;

[0047] is a proposed distribution used to generate new positions for particles.

[0048] Preferably, the input variables of the fuzzy logic control algorithm in the intelligent charging management module include the state of charge of the battery and temperature , the specific steps are as follows:

[0049] Sp1: Fuzzification of input variables: and Perform fuzzy processing, define fuzzy sets and membership functions, for , define "low", "medium", and "high" fuzzy sets, and use trigonometric functions or Gaussian functions to represent their membership. , define the fuzzy sets of “low temperature”, “appropriate temperature” and “high temperature” and the corresponding membership functions;

[0050] Sp2: Rule base establishment: Based on expert experience and experimental data, a fuzzy rule base is established. The rule form is "If yes and yes , then the charging current yes ",in , is a fuzzy set, is the charging current value;

[0051] Sp3: Fuzzy reasoning: Based on the fuzzy value of the input variable and the rule base, the fuzzy reasoning method Mamdani is used to calculate the fuzzy value of the output variable and set the input variable and Fuzzy sets and , the rule is "If yes and yes ,So yes ", then the fuzzy reasoning result is The fuzzy synthesis operation yields:

[0052]

[0053] in, represents the fuzzy reasoning result, and State of Charge and temperature The fuzzy set of Corresponds to and Combined charging current value;

[0054] Sp4: Defuzzification of output variables: Defuzzify the fuzzy value of the output variable obtained by fuzzy reasoning to obtain the accurate charging current or voltage value, and calculate the accurate output through the defuzzification method. value:

[0055]

[0056] in, Indicates the exact charging current or voltage value, Represents the fuzzy reasoning result In the A value, is the result of fuzzy reasoning In the The membership degree at each value.

[0057] Preferably, the specific steps of the constant current-constant voltage charging algorithm in the intelligent charging management module are as follows:

[0058] Sp1: Constant current charging stage: At the beginning of charging, the charging current is set to a constant value Until the battery voltage reaches the set threshold In this stage, the battery is charged quickly with a large current, and the charging time By battery capacity and charging current calculate:

[0059]

[0060] Battery level:

[0061]

[0062] Sp2: Constant voltage charging stage: When the battery voltage reaches After that, switch to constant voltage charging mode, at this time keep the charging voltage constant , the charging current gradually decreases. As the charging progresses, when the charging current decreases to the set cut-off current When the charging is finished, in the constant voltage charging stage, the charging current for:

[0063]

[0064] in, The constant voltage charging stage is at time The charging current, The charging voltage in the constant voltage charging stage, For the battery at all times The terminal voltage, For the battery at all times The internal resistance.

[0065] Preferably, in the fault diagnosis and fault-tolerant control module of the battery system, the support vector machine algorithm analyzes the operating data of the battery system through the following steps to determine whether there is a fault:

[0066] Sp1: Data preprocessing: Collecting the operating data of the battery system, including characteristic vectors such as voltage, current, and temperature , and preprocess the data to divide it into training set and test set;

[0067] Sp2: Constructing a hyperplane: In the feature space, by finding an optimal hyperplane To distinguish normal data from fault data, is the normal vector, which determines the direction of the hyperplane. is the intercept;

[0068] Sp3: Optimization solution: By solving the quadratic programming problem, determine the and , so that the interval between the two types of data is the largest, the solution formula is as follows:

[0069]

[0070] in, is the penalty parameter, is the slack variable;

[0071] The constraints are as follows:

[0072]

[0073] in, is the data label, Indicates normal, Indicates a fault. is the sample size;

[0074] Sp4: Fault diagnosis: For new input data ,pass , calculate to determine whether it is fault data.

[0075] Preferably, the model predictive control algorithm in the fault diagnosis and fault-tolerant control module of the battery system reduces the impact of the fault on the entire system through the following steps:

[0076] Sp1: System modeling: Establish a mathematical model of the battery system, expressed as a state space equation:

[0077]

[0078] in, is the system state vector, including battery voltage, current, SOC, is the control input vector, including charging current, discharging current, etc. is the output vector, including system power, voltage, is the system matrix;

[0079] Sp2: Predict future state: based on the current system state and control input sequence , predict the system status in the future:

[0080]

[0081] The prediction equation is:

[0082]

[0083]

[0084] in, Indicates at time Prediction Moment The system status, Indicates time The system status, Indicates at time The control input, Indicates at time Prediction Moment Output:

[0085] Sp3: Optimize control input: By solving the optimization problem:

[0086]

[0087] in, represents the control input sequence, represents the predicted output, , represents the weight matrix, is the control time domain, is the prediction time domain;

[0088] At the same time, the system constraints, including voltage and current limits, are met, and the optimal control input sequence is determined so that the predicted system performance indicators, including minimum power loss and maximum battery life, are optimized;

[0089] Sp4: Implement control: Input the first control obtained by optimization Applied to the system.

[0090] Preferably, the system also includes a data acquisition and transmission module for collecting battery voltage, current, temperature and other data, and transmitting these data to the battery status monitoring and prediction module, the intelligent charging management module and the battery system fault diagnosis and fault-tolerant control module.

[0091] Preferably, the system is equipped with a system monitoring and display module, which displays the battery status information in real time, including the charge state , health status , voltage, current, temperature and charging status. At the same time, the system monitoring and display module uses an intuitive graphical interface to provide historical data query and analysis functions.

[0092] Beneficial Effects

[0093] The present invention provides a UPS-specific lithium-ion battery system based on intelligent management, which has the following beneficial effects:

[0094] 1. The present invention adopts system support vector machine and random forest to establish a battery status prediction model, and combines it with particle filter algorithm for correction and optimization, which can more accurately predict the battery's SOH health state and SOC charge state, so that users can more accurately understand the battery's performance and remaining power, so as to better plan the use and maintenance of the battery. Accurate battery status monitoring and prediction helps to formulate more reasonable charging and discharging strategies. Charging and discharging according to the actual state of the battery can avoid overcharging, over-discharging and the like, thereby extending the battery's service life.

[0095] 2. The present invention adopts a system fuzzy logic control algorithm to dynamically adjust the charging parameters according to the SOC and temperature of the battery. In the initial stage of charging, a constant current charging mode is used to quickly increase the battery power. When the battery is close to full or the temperature is too high, it switches to a constant voltage charging mode. This charging method can not only replenish a large amount of power for the battery in a short time, but also ensure that the battery is full and not overcharged, while protecting the battery life. By limiting the battery temperature, the charging management module can adapt to different usage environments and conditions. When the temperature is low or high, the charging parameters can still be adjusted according to the actual state of the battery to ensure the safety and effectiveness of charging, so that the UPS system can operate stably in various complex environments.

[0096] 3. The present invention adopts a system that uses a support vector machine algorithm to analyze the operating data of the battery system, and can detect faults in a timely manner. When a fault is detected, the model predictive control algorithm reduces the impact of the fault on the entire system by optimizing the control input, such as adjusting the charging current and the discharging current, to ensure that the UPS system can still operate normally in the event of a battery failure. Through fault diagnosis and fault-tolerant control, potential problems of the battery system can be discovered in advance, and corresponding measures can be taken to deal with them to avoid further expansion of the fault. This not only improves the reliability of the UPS system, but also reduces the risk of safety accidents caused by battery failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0098] Figure 2 It is a diagram of the system operation steps of the present invention;

[0099] Figure 3 It is a schematic diagram of system data transmission of the present invention. DETAILED DESCRIPTION

[0100] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment one:

[0102] like Figures 1 to 3 As shown, the UPS-specific lithium-ion battery system based on intelligent management includes a battery status monitoring and prediction module, an intelligent charging management module, and a battery system fault diagnosis and fault-tolerant control module, wherein:

[0103] Battery status monitoring and prediction module: First, the support vector machine algorithm and random forest algorithm are used to establish a battery status prediction model. According to the input real-time data, including voltage, current, and temperature, the battery status is predicted. and , the prediction results are used as the input of the particle filter algorithm. The particle filter algorithm corrects and optimizes the prediction results by continuously updating the position and weight of the particles. The support vector machine and random forest establish a battery state prediction model, and are combined with the particle filter algorithm for correction and optimization. It can more accurately predict the battery's SOH health state and SOC charge state, so that users can more accurately understand the battery's performance and remaining power, so as to better plan the use and maintenance of the battery. For example, in a UPS system, accurate SOC prediction can ensure that the battery can provide sufficient power support during a power outage and avoid sudden power outages in the system due to power estimation errors. Accurate battery state monitoring and prediction help to formulate more reasonable charging and discharging strategies. Charging and discharging according to the actual state of the battery can avoid overcharging, over-discharging, etc., thereby extending the battery life. For example, when it is predicted that the battery is close to full, the intelligent charging management module will reduce the charging current and switch to constant voltage charging mode to protect the battery life.

[0104] Intelligent charging management module: The fuzzy logic control algorithm dynamically adjusts the charging parameters according to the current state SOC and temperature T of the battery. In the early stage of charging, when the SOC is low and the battery temperature is within the appropriate range, the fuzzy logic controller outputs a larger charging current value and uses the constant current charging mode to quickly increase the battery power. As charging progresses, when the battery is close to being fully charged and the SOC is high or the temperature is too high, the fuzzy logic controller will reduce the charging current and switch to the constant voltage charging mode to ensure that the battery is fully charged and not overcharged, while protecting the battery life. The fuzzy logic control algorithm dynamically adjusts the charging parameters according to the SOC and temperature of the battery. In the early stage of charging, the constant current charging mode is used to quickly increase the battery power. When the battery is nearly full or the temperature is too high, it switches to constant voltage charging mode. This charging method can not only replenish a large amount of power for the battery in a short time, but also ensure that the battery is fully charged and not overcharged, while protecting the battery life. For example, in some scenarios where charging time is required, such as emergency backup power supply systems, the charging management module can complete charging as quickly as possible while ensuring battery safety. By limiting the battery temperature, the charging management module can adapt to different use environments and conditions. When the temperature is low or high, it can still adjust the charging parameters according to the actual state of the battery to ensure the safety and effectiveness of charging, so that the UPS system can operate stably in various complex environments.

[0105] Fault diagnosis and fault-tolerant control module of the battery system: The support vector machine algorithm is used to analyze the operating data of the battery system to determine whether there is a fault. When a fault is detected, the model predictive control algorithm predicts the future performance based on the mathematical model of the battery system, and optimizes the control input, such as adjusting the charging current and discharging current, to reduce the impact of the fault on the entire system, ensuring that the UPS system can still operate normally in the event of a battery failure. At the same time, the MPC algorithm dynamically adjusts the control strategy according to the development of the fault, and uses the support vector machine algorithm to analyze the operating data of the battery system to detect the fault in time. When a fault is detected, the model predictive control algorithm optimizes the control input, such as adjusting the charging current and discharging current, to reduce the impact of the fault on the entire system, ensuring that the UPS system can still operate normally in the event of a battery failure. Through fault diagnosis and fault-tolerant control, potential problems of the battery system can be discovered in advance, and corresponding measures can be taken to deal with them to avoid further expansion of the fault, which not only improves the reliability of the UPS system, but also reduces the risk of safety accidents caused by battery failure.

[0106] The system also includes a data acquisition and transmission module, which is used to collect data such as battery voltage, current, temperature, etc., and transmit these data to the battery status monitoring and prediction module, the intelligent charging management module and the fault diagnosis and fault-tolerant control module of the battery system. The data acquisition and transmission module collects data such as battery voltage, current, temperature, etc., and transmits them to each module. The system monitoring and display module displays the battery status information in real time, including the state of charge SOC, the state of health SOH, voltage, current, temperature and charging status. Users can understand the battery operation status at any time through the intuitive graphical interface, detect abnormal conditions in time and take measures.

[0107] The system is equipped with a system monitoring and display module, which displays the battery status information in real time, including the charge status , health status , voltage, current, temperature and charging status. At the same time, the system monitoring and display module uses an intuitive graphical interface to provide historical data query and analysis functions. Users can understand the battery usage trend and performance changes through the analysis of historical data, and provide a basis for battery maintenance and replacement. For example, by analyzing the battery's SOH change trend, the end of the battery life can be predicted, and replacement preparations can be made in advance to avoid system failures due to battery aging.

[0108] The entire system structure uses support vector machines, random forests and particle filter algorithms to accurately predict the battery's SOH and SOC, providing a basis for reasonable charging and discharging and extending battery life; the intelligent charging management module uses fuzzy logic control algorithms to dynamically adjust charging parameters according to battery SOC and temperature to achieve a balance between fast charging and battery protection, and can adapt to different environments; the fault diagnosis and fault-tolerant control module uses support vector machine algorithms to detect faults in a timely manner, and the model predictive control algorithm optimizes control inputs in the event of a fault, ensuring that the UPS system continues to operate normally under battery failure; in addition, the data acquisition and transmission module and the system monitoring and display module can monitor the battery status in real time and provide historical data query and analysis, making it convenient for users to understand the battery status and make maintenance decisions. Specific embodiment 2:

[0110] like Figures 1 to 3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed:

[0111] The support vector machine algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps:

[0112] Sp1: Collect a large amount of operating data of UPS-specific lithium-ion batteries under different usage conditions, including voltage , Current ,temperature , Charge and discharge cycle times , and mark these data as the training set:

[0113]

[0114] in, represents the input feature vector, Indicates the corresponding battery status label, including or value;

[0115] Sp2: Determine the optimal hyperplane under constraints by solving the quadratic programming problem , the solution formula is as follows:

[0116]

[0117] The constraints are as follows:

[0118]

[0119] in, is the weight vector, is the bias term, is the penalty parameter used to balance the penalty for interval maximization and misclassification, is the slack variable, Representation vector The norm of , is the number of training samples;

[0120] Sp3: For new input data ,pass Calculate the predicted battery state, represents a symbolic function, if ,but ,like ,but .

[0121] The random forest algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps:

[0122] Sp1: Randomly extract multiple sample subsets from the original training data, and the size of each subset is the same as the original training set.

[0123] Sp2: For each sample subset, a decision tree is constructed. When constructing the decision tree, some features are randomly selected as split features at each node, and the optimal split point is selected for splitting;

[0124] Sp3: For new input data, the prediction results of multiple decision trees are integrated and the final prediction result is determined by majority voting, that is:

[0125]

[0126] in, For the final prediction result, Represents input data, is the number of decision trees, For the A decision tree, Indicates majority.

[0127] The particle filter algorithm in the battery status monitoring and prediction module corrects and optimizes the prediction results through the following steps:

[0128] Sp1: Initialization: At the initial moment ,generate Particles , and according to the prior distribution Determine the initial weight of each particle ;

[0129] Sp2: Prediction: Based on the system state transition model , predict the position of each particle at the current moment ,in, is the control input;

[0130] Sp3: Update: Based on the observation model and observations , calculate the observation likelihood of each particle , then update the particle weights ;

[0131] Sp4: Resampling: If the effective number of samples of a particle Less than the set threshold , then resample to obtain a new particle set , and reset the weights to ,in, represents the number of particles, Indicates at time No. The weight of each particle;

[0132] Sp5: State estimation: According to the weight and position of the particle, by calculating the moment An estimate of the system state To estimate the system state:

[0133]

[0134] The particle weight update formula is:

[0135]

[0136] in,

[0137] , They are Moment The weight and position of each particle;

[0138] , They are Moment and Moment The weight of each particle;

[0139] At a given particle position In the case of The conditional probability of

[0140] According to the system state transition model, from time The particle position Transfer to the particle position at time probability;

[0141] is a proposed distribution used to generate new positions for particles.

[0142] The input variables of the fuzzy logic control algorithm in the intelligent charging management module include the battery state of charge and temperature , the specific steps are as follows:

[0143] Sp1: Fuzzification of input variables: and Perform fuzzy processing, define fuzzy sets and membership functions, for , define "low", "medium", and "high" fuzzy sets, and use trigonometric functions or Gaussian functions to represent their membership. , define the fuzzy sets of “low temperature”, “appropriate temperature” and “high temperature” and the corresponding membership functions;

[0144] Sp2: Rule base establishment: Based on expert experience and experimental data, a fuzzy rule base is established. The rule form is "If yes and yes , then the charging current yes ",in , is a fuzzy set, is the charging current value;

[0145] Sp3: Fuzzy reasoning: Based on the fuzzy value of the input variable and the rule base, the fuzzy reasoning method Mamdani is used to calculate the fuzzy value of the output variable and set the input variable and Fuzzy sets and , the rule is "If yes and yes ,So yes ", then the fuzzy reasoning result is The fuzzy synthesis operation yields:

[0146]

[0147] in, represents the fuzzy reasoning result, and State of Charge and temperature The fuzzy set of Corresponds to and Combined charging current value;

[0148] Sp4: Defuzzification of output variables: Defuzzify the fuzzy value of the output variable obtained by fuzzy reasoning to obtain the accurate charging current or voltage value. The accurate output is calculated by defuzzification method. value:

[0149]

[0150] in, Indicates the exact charging current or voltage value, Represents the fuzzy reasoning result In the A value, is the result of fuzzy reasoning In the The membership degree at each value.

[0151] The specific steps of the constant current-constant voltage charging algorithm in the intelligent charging management module are as follows:

[0152] Sp1: Constant current charging stage: At the beginning of charging, the charging current is set to a constant value Until the battery voltage reaches the set threshold In this stage, the battery is charged quickly with a large current, and the charging time By battery capacity and charging current calculate:

[0153]

[0154] Battery level:

[0155]

[0156] Sp2: Constant voltage charging stage: When the battery voltage reaches After that, switch to constant voltage charging mode, at this time keep the charging voltage constant , the charging current gradually decreases. As the charging progresses, when the charging current decreases to the set cut-off current When the charging is finished, in the constant voltage charging stage, the charging current for:

[0157]

[0158] in, The constant voltage charging stage is at time The charging current, The charging voltage in the constant voltage charging stage, For the battery at all times The terminal voltage, For the battery at all times The internal resistance.

[0159] In the fault diagnosis and fault-tolerant control module of the battery system, the support vector machine algorithm analyzes the operating data of the battery system through the following steps to determine whether there is a fault:

[0160] Sp1: Data preprocessing: Collecting the operating data of the battery system, including characteristic vectors such as voltage, current, and temperature , and preprocess the data to divide it into training set and test set;

[0161] Sp2: Constructing a hyperplane: In the feature space, by finding an optimal hyperplane To distinguish normal data from fault data, is the normal vector, which determines the direction of the hyperplane. is the intercept;

[0162] Sp3: Optimization solution: By solving the quadratic programming problem, determine the and , so that the interval between the two types of data is the largest, the solution formula is as follows:

[0163]

[0164] in, is the penalty parameter, is the slack variable;

[0165] The constraints are as follows:

[0166]

[0167] in, is the data label, Indicates normal, Indicates a fault. is the sample size;

[0168] Sp4: Fault diagnosis: For new input data ,pass , calculate to determine whether it is fault data.

[0169] The model predictive control algorithm in the fault diagnosis and fault-tolerant control module of the battery system reduces the impact of the fault on the entire system through the following steps:

[0170] Sp1: System modeling: Establish a mathematical model of the battery system, expressed as a state space equation:

[0171]

[0172] in, is the system state vector, including battery voltage, current, SOC, is the control input vector, including charging current, discharging current, etc. is the output vector, including system power, voltage, is the system matrix;

[0173] Sp2: Predict future state: based on the current system state and control input sequence , predict the system status in the future:

[0174]

[0175] The prediction equation is:

[0176]

[0177]

[0178] in, Indicates at time Prediction Moment The system status, Indicates time The system status, Indicates at time The control input, Indicates at time Prediction Moment Output:

[0179] Sp3: Optimize control input: By solving the optimization problem:

[0180]

[0181] in, represents the control input sequence, represents the predicted output, , represents the weight matrix, is the control time domain, is the prediction time domain;

[0182] At the same time, the system constraints, including voltage and current limits, are met, and the optimal control input sequence is determined so that the predicted system performance indicators, including minimum power loss and maximum battery life, are optimized;

[0183] Sp4: Implement control: Input the first control obtained by optimization Applied to the system. Specific embodiment three:

[0185] like Figures 1 to 3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed:

[0186] The data processing content of the entire system is as follows:

[0187] The data acquisition and transmission module collects battery voltage, current, temperature and other data, and transmits the collected data to the battery status monitoring and prediction module, the intelligent charging management module and the battery system fault diagnosis and fault-tolerant control module, which becomes the basic data for analysis and processing by other modules.

[0188] Battery status monitoring and prediction module: receives voltage, current, and temperature data from the data acquisition and transmission module. The operating logic of the battery status monitoring and prediction module is as follows:

[0189] First, support vector machine and random forest are used to establish a battery status prediction model. The support vector machine algorithm collects a large amount of operating data of UPS-specific lithium-ion batteries under different usage conditions, including voltage , Current ,temperature , Charge and discharge cycle times , as the training set, the optimal hyperplane is determined by solving the quadratic programming problem, and then the battery status is predicted for the new input data; the random forest algorithm randomly extracts multiple sample subsets from the original training data to construct a decision tree, and the prediction results of multiple decision trees are combined and the final prediction result is determined by majority voting;

[0190] Then, the prediction results of support vector machine and random forest are used as the input of particle filter algorithm. The particle filter algorithm generates particles by initialization and determines the initial weights, predicts the particle positions according to the system state transition model, calculates the particle observation likelihood according to the observation model and the observed value and updates the weights, resamples when necessary, and finally corrects and optimizes the prediction results by calculating the estimated value of the system state.

[0191] The data movement of the battery status monitoring and prediction module is as follows: the predicted battery SOH and SOC data are provided to the intelligent charging management module on the one hand, and to the system monitoring and display module on the other hand for display;

[0192] The data of the battery status monitoring and prediction module goes to: the intelligent charging management module and the system monitoring and display module.

[0193] Intelligent charging management module: obtains the battery SOC data from the battery status monitoring and prediction module, and obtains the battery temperature T data from the data acquisition and transmission module. The operation logic of the intelligent charging management module is as follows:

[0194] The fuzzy logic control algorithm fuzzifies SOC and T, defines fuzzy sets and membership functions, establishes a fuzzy rule base, calculates the fuzzy value of the output variable according to the fuzzy reasoning method, and finally performs defuzzification to obtain the accurate charging current or voltage value;

[0195] During the charging process, the constant current-constant voltage charging algorithm performs charging at different stages according to the battery status. At the beginning of charging, it is constant current charging. When the battery voltage reaches the set threshold, it switches to constant voltage charging. In the constant voltage charging stage, the charging current is calculated according to the battery terminal voltage and internal resistance. Charging ends when the charging current decreases to the set cut-off current.

[0196] The data movement of the intelligent charging management module is as follows: charging the battery according to the calculated charging current and voltage parameters, and feeding back the charging status and other data to the system monitoring and display module;

[0197] The data of the intelligent charging management module goes to: battery charging, system monitoring and display module.

[0198] Fault diagnosis and fault-tolerant control module of the battery system: Receives battery operation data from the data acquisition and transmission module, including voltage, current, temperature, etc. Its operation logic is as follows:

[0199] The support vector machine algorithm pre-processes the data, divides it into training set and test set, constructs a hyperplane, determines the normal vector and intercept of the hyperplane by solving the quadratic programming problem, and then determines whether there is a fault in the new input data;

[0200] When a fault is detected, the model predictive control algorithm establishes a mathematical model of the battery system, a state-space equation, and predicts the system state in the future based on the current system state and the control input sequence. The optimal control input sequence is determined by solving the optimization problem, such as adjusting the charging current and the discharging current, and the first optimized control input is applied to the system to reduce the impact of the fault on the entire system.

[0201] Data movement of the fault diagnosis and fault-tolerant control module of the battery system: Feedback of fault diagnosis results and fault-tolerant control strategies to the system monitoring and display module;

[0202] The data destination of the battery system's fault diagnosis and fault-tolerant control module: system monitoring and display module.

[0203] System monitoring and display module: obtains the battery's SOH, SOC, voltage, current, and temperature data from the battery status monitoring and prediction module, obtains charging status data from the intelligent charging management module, and obtains fault diagnosis results and fault-tolerant control strategy data from the battery system's fault diagnosis and fault-tolerant control module. Its operating logic is as follows:

[0204] Real-time display of battery status information, including state of charge (SOC), state of health (SOH), voltage, current, temperature, and charging status. An intuitive graphical interface is also used to provide historical data query and analysis functions.

[0205] Data trends of the fault diagnosis and fault-tolerant control module of the battery system: display data to users for query and analysis;

[0206] The data destination of the battery system's fault diagnosis and fault-tolerant control module: the user. Specific embodiment four:

[0208] like Figures 1 to 3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed:

[0209] The operation steps of the whole system are as follows:

[0210] Sp1: The data acquisition and transmission module collects battery voltage, current, temperature and other data and transmits them to other modules;

[0211] Sp2: The battery status monitoring and prediction module uses support vector machine, random forest and particle filter algorithms to predict the battery's SOH and SOC based on input data;

[0212] Sp3: The intelligent charging management module determines the charging parameters and charges the battery according to the SOC and temperature of the battery through the fuzzy logic control algorithm and constant current-constant voltage charging algorithm;

[0213] Sp4: The fault diagnosis and fault-tolerant control module of the battery system uses the support vector machine algorithm to detect faults and uses the model predictive control algorithm to adjust the control input in the event of a fault to reduce the impact of the fault;

[0214] Sp5: The system monitoring and display module displays battery status information in real time and provides historical data query and analysis functions.

[0215] The entire system structure realizes accurate monitoring and prediction of battery status, provides a basis for the rational use and maintenance of batteries, and prolongs battery life. Intelligent charging management ensures fast and safe charging of batteries while protecting battery performance. Effective fault diagnosis and fault-tolerant control improve the reliability and stability of the UPS system in the event of battery failure. The system monitoring and display module allows users to understand the battery status at any time, making it easier to make maintenance and management decisions. Specific embodiment five:

[0217] like Figures 1 to 3 As shown, according to the contents in the above specific embodiments, the following contents are further disclosed:

[0218] In actual use, each module of the entire system has at least the following hardware content:

[0219] The data acquisition and transmission module has sensors, signal conditioning circuits, analog-to-digital converters ADC, microcontrollers MCU or microprocessors MPU, communication interfaces, etc. in its actual hardware structure. The specific contents are as follows:

[0220] sensor:

[0221] Voltage sensor: used to measure the terminal voltage of the battery, usually using a voltage divider resistor or Hall effect voltage sensor, etc. These sensors convert the high voltage of the battery into a low voltage signal suitable for subsequent processing. For example, a resistor voltage divider voltage sensor is used to connect high-precision resistors in series to proportionally reduce the battery voltage according to the resistance ratio, and then the divided voltage signal is input into the subsequent signal conditioning circuit;

[0222] Current sensor: Hall current sensor is generally used to measure the charge and discharge current of the battery. Hall current sensor is based on the Hall effect and can measure current without contact. It has the advantages of high accuracy and fast response speed. When current passes through the sensor, it will generate a magnetic field proportional to the current. The sensor detects the magnetic field strength and converts it into an electrical signal to output the corresponding current measurement value.

[0223] Temperature sensor: such as thermocouples or thermistors, used to monitor the temperature of the battery. Thermocouples measure temperature based on the principle that a circuit composed of two different metals generates thermoelectric potential under a temperature gradient; thermistors measure temperature based on the change in their resistance value with temperature. Temperature sensors are usually installed on the surface of the battery or near the battery to accurately reflect the battery temperature.

[0224] Signal conditioning circuit: Filter, amplify and linearize the voltage, current and temperature signals collected by the sensor. Filtering can remove noise and interference in the signal, amplification can increase the weak signal to a suitable amplitude, and linearization can convert the nonlinear output of the sensor into a linear signal for subsequent analog-to-digital conversion and processing. For example, a low-pass filter is used to remove high-frequency noise, and an operational amplifier is used to amplify the signal.

[0225] Analog-to-digital converter (ADC): Converts the conditioned analog signal into a digital signal. The resolution and sampling rate of the ADC affect the precision and accuracy of the data. A higher resolution can provide more precise measurements, while a suitable sampling rate can ensure accurate capture of changes in the battery status. For example, a 12-bit or 16-bit ADC can be selected, and the sampling rate can be determined based on the system's requirements for data acquisition speed.

[0226] Microcontroller (MCU) or microprocessor (MPU): responsible for controlling the data acquisition process, including setting the ADC's operating parameters, reading the converted digital data, and transmitting the data to other modules through the communication interface. The MCU or MPU can also perform some preliminary data processing and packaging, such as calculating statistical information such as the average and standard deviation. It can be a single-chip microcomputer designed specifically for industrial control, or a more powerful embedded processor;

[0227] Communication interface: used to transmit the collected data to the battery status monitoring and prediction module, the intelligent charging management module and the fault diagnosis and fault-tolerant control module of the battery system. Common communication interfaces include CAN bus, RS-485, Ethernet, etc. The CAN bus has the characteristics of high reliability and strong real-time performance, and is suitable for data transmission in industrial environments; RS-485 has the advantages of low cost and long transmission distance; Ethernet is suitable for systems that require high-speed data transmission and networking functions.

[0228] The battery status monitoring and prediction module has a processor, memory, and algorithm implementation circuit structure in actual hardware structure. The specific contents are as follows:

[0229] Processor: It can be a high-performance digital signal processor (DSP) or field programmable gate array (FPGA). DSP is good at digital signal processing and algorithm operation, and can quickly execute algorithms such as support vector machine, random forest and particle filter. FPGA has parallel processing capability and reconfigurability, and can customize hardware logic as needed to improve the execution efficiency of the algorithm. For example, if a DSP chip is used, it can implement complex mathematical operations and data processing through programming, receive digital data from the data acquisition and transmission module, and calculate and optimize the battery state prediction model.

[0230] Memory: includes random access memory (RAM) and read-only memory (ROM). RAM is used to store temporary data during the algorithm operation, such as the position and weight of particles; ROM is used to store program code and pre-trained model parameters. For example, high-speed static RAM is used to store real-time calculated data, and flash memory is used as ROM to store solidified programs and model parameters.

[0231] Algorithm implementation circuit: For some key algorithm operations, the calculation can be accelerated by specially designed hardware circuits. For example, for some matrix operations and feature extraction in support vector machines and random forests, dedicated digital signal processing circuits or ASICs (application-specific integrated circuits) can be used to increase the calculation speed.

[0232] The actual hardware structure of the intelligent charging management module includes a charging controller, a power conversion circuit, and a temperature monitoring and protection circuit. The specific contents are as follows:

[0233] Charging controller: This is the core hardware of the intelligent charging management module. The charging controller is usually composed of analog and digital circuits, and can adjust the charging current and voltage according to the input control signal (output from the fuzzy logic control algorithm). It can use a dedicated charging control chip, such as the battery charging management chip produced by Texas Instruments (TI) and other companies. The charging controller realizes the switching of constant current-constant voltage charging mode by controlling the on and off time of power switching devices (such as MOSFET or IGBT). In the constant current charging stage, the controller keeps the charging current constant; in the constant voltage charging stage, the controller adjusts the power switching device to keep the charging voltage constant, and dynamically adjusts the charging current according to the battery status;

[0234] Power conversion circuit: including DC-DC converters, etc. DC-DC converters convert the input power voltage into a voltage and current suitable for battery charging. For example, a step-down (Buck) DC-DC converter is used to reduce a higher input voltage to a voltage suitable for battery charging. The inductor, capacitor, power switch device and other components in the power conversion circuit work together to achieve voltage and current conversion;

[0235] Temperature monitoring and protection circuit: connected to the temperature sensor to monitor the battery temperature in real time. When the temperature is too high or too low, the temperature monitoring and protection circuit will send a signal to the charging controller, and the charging controller will adjust the charging parameters or suspend charging accordingly to protect the battery. For example, a comparator circuit is used to compare the output of the temperature sensor with a preset temperature threshold. When the temperature exceeds the threshold, the protection action is triggered.

[0236] The fault diagnosis and fault-tolerant control module of the battery system has a fault diagnosis circuit, a fault-tolerant control circuit, a monitoring function and an alarm circuit in the actual hardware structure. The specific contents are as follows:

[0237] Fault diagnosis circuit: It consists of a sensor signal processing circuit and a fault judgment logic circuit. The sensor signal processing circuit further analyzes and processes the voltage, current, temperature and other signals from the data acquisition and transmission module to extract fault features. The fault judgment logic circuit determines whether a fault exists based on the judgment result of the support vector machine algorithm. For example, by calculating the mean, variance, spectrum and other features of the signal and comparing them with the feature values ​​under normal conditions, when the deviation exceeds a certain threshold, it is considered that a fault may exist.

[0238] Fault-tolerant control circuit: When a fault is detected, the fault-tolerant control circuit adjusts control parameters such as charging current and discharging current according to the output of the model predictive control algorithm. The fault-tolerant control circuit can adjust the control parameters through devices such as digital potentiometers and analog multipliers. For example, a digital potentiometer can change the resistance value to adjust the current in the charging or discharging circuit; an analog multiplier can adjust the output voltage or current value according to the input control signal;

[0239] Monitoring and alarm circuit: monitors the operating status of the battery system in real time and sends out an alarm signal when a fault is detected. The alarm signal can be an audible and visual alarm, a fault code sent to the monitoring system through the communication interface, etc. The monitoring and alarm circuit can also record the time and type of the fault for subsequent fault analysis and processing.

[0240] The system monitoring and display module has a display screen, a microcontroller or a microprocessor, a storage device and an input device in terms of actual hardware structure. The specific contents are as follows:

[0241] Display screen: It can be a liquid crystal display (LCD) or an organic light-emitting diode display (OLED), etc. The display screen is used to display battery status information, such as state of charge SOC, state of health SOH, voltage, current, temperature, and charging status. The size and resolution of the display screen are determined according to actual application requirements. It is usually required to have a clear display effect and a good viewing angle;

[0242] Microcontroller or microprocessor: Responsible for receiving data from other modules and processing it into a format suitable for display on the display screen. At the same time, it also handles user input operations, such as querying historical data, etc. Similar to the microcontrollers or microprocessors in other modules, it can perform functions such as data calculation, storage management and communication control;

[0243] Storage device: used to store historical data for users to query and analyze. The storage device can be a flash memory chip, SD card, etc. When the user queries the historical data, the microcontroller or microprocessor reads the corresponding data from the storage device and displays it on the display screen;

[0244] Input devices: such as buttons and touch screens, are used for users to interact with the system. Users can use input devices to set display parameters, query historical data for a specific time period, and perform other operations.

[0245] The hardware structure and operation logic of the entire system are as follows:

[0246] The data acquisition and transmission module collects data such as battery voltage, current and temperature through sensors. After signal conditioning and analog-to-digital conversion, the microcontroller or microprocessor sends the data to the battery status monitoring and prediction module through the communication interface. The processor of the battery status monitoring and prediction module uses the algorithm and model parameters in the memory to process and analyze the data, predict the battery's SOH and SOC, and send the results to the intelligent charging management module and the system monitoring and display module;

[0247] The charging controller of the intelligent charging management module controls the power conversion circuit to achieve constant current-constant voltage charging according to the SOC and temperature of the battery, combined with the output of the fuzzy logic control algorithm. At the same time, the temperature monitoring and protection circuit monitors the temperature in real time to ensure charging safety.

[0248] The fault diagnosis circuit of the fault-tolerant control module of the battery system continuously analyzes the data transmitted by the data acquisition and transmission module to determine whether there is a fault. Once a fault is detected, the fault-tolerant control circuit adjusts the control parameters according to the model prediction control algorithm, and the monitoring and alarm circuit sends an alarm signal;

[0249] The system monitoring and display module receives data from other modules, processes it through a microcontroller or microprocessor, and displays it on the display screen. Users can operate it through input devices, query historical data, etc.

[0250] In short, the hardware of each module cooperates with each other to realize intelligent management of the UPS dedicated lithium-ion battery system through data collection, transmission, processing and control, ensure the safe and efficient operation of the battery, and provide users with an intuitive monitoring and operation interface.

[0251] It should 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 statement "comprising a reference structure" do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0252] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. UPS-specific lithium-ion battery system based on intelligent management, characterized by: The system includes a battery status monitoring and prediction module, an intelligent charging management module and a battery system fault diagnosis and fault tolerance control module, wherein: The battery state monitoring and prediction module: firstly, a battery state prediction model is established by using a support vector machine algorithm and a random forest algorithm, and the SOH and SOC of the battery are predicted according to the input real-time data, including voltage, current, and temperature, and the prediction results are used as inputs of a particle filter algorithm, which corrects and optimizes the prediction results by continuously updating the positions and weights of particles; The intelligent charging management module: the fuzzy logic control algorithm dynamically adjusts the charging parameters according to the current state SOC and temperature T of the battery. In the early stage of charging, when the SOC is low and the battery temperature is within a suitable range, the fuzzy logic controller outputs a larger charging current value and adopts a constant current charging mode to quickly increase the battery power. As the charging progresses, when the battery is close to being fully charged and the SOC is high or the temperature is too high, the fuzzy logic controller will reduce the charging current and switch to a constant voltage charging mode to ensure that the battery is fully charged and not overcharged, while protecting the battery life; The fault diagnosis and fault-tolerant control module of the battery system uses a support vector machine algorithm to analyze the operating data of the battery system to determine whether there is a fault. When a fault is detected, the model predictive control algorithm predicts the future performance based on the mathematical model of the battery system, and optimizes the control input, such as adjusting the charging current and discharging current, to reduce the impact of the fault on the entire system, ensuring that the UPS system can still operate normally in the event of a battery failure. At the same time, the MPC algorithm dynamically adjusts the control strategy according to the development of the fault.

2. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The support vector machine algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps: Sp1: Collect a large amount of operating data of UPS-specific lithium-ion batteries under different usage conditions, including voltage , Current ,temperature , Charge and discharge cycle times , and mark these data as the training set: ； in, represents the input feature vector, Indicates the corresponding battery status label, including or value; Sp2: Determine the optimal hyperplane under constraints by solving the quadratic programming problem , the solution formula is as follows: ； The constraints are as follows: ； in, is the weight vector, is the bias term, is the penalty parameter used to balance the penalty for interval maximization and misclassification, is the slack variable, Representation vector The norm of , is the number of training samples; Sp3: For new input data ,pass Calculate the predicted battery state, represents a symbolic function, if ,but ,like ,but .

3. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The random forest algorithm in the battery status monitoring and prediction module establishes a battery status prediction model through the following steps: Sp1: Randomly extract multiple sample subsets from the original training data, and the size of each subset is the same as the original training set; Sp2: For each sample subset, a decision tree is constructed. When constructing the decision tree, some features are randomly selected as split features at each node, and the optimal split point is selected for splitting; Sp3: For new input data, the prediction results of multiple decision trees are integrated and the final prediction result is determined by majority voting, that is: ； in, For the final prediction result, Represents input data, is the number of decision trees, For the A decision tree, Indicates majority.

4. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The particle filter algorithm in the battery status monitoring and prediction module corrects and optimizes the prediction results through the following steps: Sp1: Initialization: At the initial moment ,generate Particles , and according to the prior distribution Determine the initial weight of each particle ; Sp2: Prediction: Based on the system state transition model , predict the position of each particle at the current moment ,in, is the control input; Sp3: Update: Based on the observation model and observations , calculate the observation likelihood of each particle , then update the particle weights ; Sp4: Resampling: If the effective number of samples of a particle Less than the set threshold , then resample to obtain a new particle set , and reset the weights to ,in, represents the number of particles, Indicates at time No. The weight of each particle; Sp5: State estimation: According to the weight and position of the particle, by calculating the moment An estimate of the system state To estimate the system state: ； The particle weight update formula is: ； in, , They are Moment The weight and position of each particle; , They are Moment and Moment The weight of each particle; At a given particle position In the case of The conditional probability of According to the system state transition model, from time The particle position Transfer to the particle position at time probability; is a proposed distribution used to generate new positions for particles.

5. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The input variables of the fuzzy logic control algorithm in the intelligent charging management module include the battery state of charge and temperature , the specific steps are as follows: Sp1: Fuzzification of input variables: and Perform fuzzy processing, define fuzzy sets and membership functions, for , define "low", "medium", and "high" fuzzy sets, and use trigonometric functions or Gaussian functions to represent their membership. , define the fuzzy sets of "low temperature", "appropriate temperature", "high temperature" and the corresponding membership functions; Sp2: Rule base establishment: Based on expert experience and experimental data, a fuzzy rule base is established. The rule form is "If yes and yes , then the charging current yes ",in , is a fuzzy set, is the charging current value; Sp3: Fuzzy reasoning: Based on the fuzzy value of the input variable and the rule base, the fuzzy reasoning method Mamdani is used to calculate the fuzzy value of the output variable and set the input variable and Fuzzy sets and , the rule is "if yes and yes ,So yes ", then the fuzzy reasoning result is The fuzzy synthesis operation yields: ； in, represents the fuzzy reasoning result, and State of Charge and temperature The fuzzy set of Corresponds to and Combined charging current value; Sp4: Defuzzification of output variables: Defuzzify the fuzzy value of the output variable obtained by fuzzy reasoning to obtain the accurate charging current or voltage value, and calculate the accurate output through the defuzzification method. value: ； in, Indicates the exact charging current or voltage value, Represents the fuzzy reasoning result In the A value, is the result of fuzzy reasoning In the The membership degree at each value.

6. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The specific steps of the constant current-constant voltage charging algorithm in the intelligent charging management module are as follows: Sp1: Constant current charging stage: At the beginning of charging, the charging current is set to a constant value , until the battery voltage reaches the set threshold In this stage, the battery is charged quickly with a large current, and the charging time By battery capacity and charging current calculation : ； Battery level: ； Sp2: Constant voltage charging stage: When the battery voltage reaches After that, switch to constant voltage charging mode, at this time keep the charging voltage constant , the charging current gradually decreases. As the charging progresses, when the charging current decreases to the set cut-off current When the charging is finished, in the constant voltage charging stage, the charging current for: ； in, During the constant voltage charging phase, The charging current, The charging voltage in the constant voltage charging stage, For the battery at all times The terminal voltage, For the battery at all times The internal resistance.

7. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: In the fault diagnosis and fault-tolerant control module of the battery system, the support vector machine algorithm analyzes the operating data of the battery system through the following steps to determine whether there is a fault: Sp1: Data preprocessing: Collect the operating data of the battery system, including characteristic vectors such as voltage, current, and temperature , and preprocess the data to divide it into training set and test set; Sp2: Constructing a hyperplane: In the feature space, by finding an optimal hyperplane To distinguish normal data from fault data, is the normal vector, which determines the direction of the hyperplane. is the intercept; Sp3: Optimization solution: By solving the quadratic programming problem, determine the and , so that the interval between the two types of data is the largest, the solution formula is as follows: ； in, is the penalty parameter, is the slack variable; The constraints are as follows: ； in, is the data label, Indicates normal, Indicates a fault. is the sample size; Sp4: Fault diagnosis: For new input data ,pass , calculate to determine whether it is fault data.

8. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1 is characterized in that: The model predictive control algorithm in the fault diagnosis and fault-tolerant control module of the battery system reduces the impact of the fault on the entire system through the following steps: Sp1: System modeling: Establish a mathematical model of the battery system, expressed as a state space equation: ； in, is the system state vector, including battery voltage, current, SOC, is the control input vector, including charging current, discharging current, etc. is the output vector, including system power and voltage, is the system matrix; Sp2: Predict future state: based on the current system state and control input sequence , predict the system status in the future: ； The prediction equation is: ； ； in, Indicates at time Prediction Moment The system status, Indicates time The system status, Indicates at time The control input, Indicates at time Prediction Moment Output: Sp3: Optimize control input: By solving the optimization problem: ； in, represents the control input sequence, represents the predicted output, , represents the weight matrix, is the control time domain, is the prediction time domain; At the same time, the system constraints, including voltage and current limits, are met, and the optimal control input sequence is determined so that the predicted system performance indicators, including minimum power loss and maximum battery life, are optimized; Sp4: Implement control: Input the first control obtained by optimization Applied to the system.

9. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1, characterized in that: The system also includes a data acquisition and transmission module for collecting battery voltage, current, temperature and other data, and transmitting the data to a battery status monitoring and prediction module, an intelligent charging management module and a battery system fault diagnosis and fault tolerance control module.

10. The UPS dedicated lithium-ion battery system based on intelligent management according to claim 1, characterized in that: The system is equipped with a system monitoring and display module, which displays the battery status information in real time, including the charge status , health status , voltage, current, temperature and charging status. At the same time, the system monitoring and display module uses an intuitive graphical interface to provide historical data query and analysis functions.

Images