Lithium battery integrated management system based on Internet of Things and block chain technology

Abstract

The invention provides a lithium battery comprehensive management system based on the Internet of Things and a block chain technology, and relates to the technical field of lithium battery comprehensive management. Comprising a data acquisition module, a data transmission module, an edge cloud cooperative processing module, an intelligent contract management module, a data security protection module, a cross-domain fusion module, a fault prediction module, an optimization scheduling module and a knowledge updating module. The system overcomes the limitation of the conventional sensor configuration, and can capture the operation information of the lithium battery in all directions. Basic electrical parameters of the battery can be monitored in real time, environmental factors are taken into consideration, a rich and accurate data basis is provided for accurately mastering the real state of the battery, management decision errors caused by one-sided data are effectively avoided, and the reliability and scientificity of battery management are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium battery integrated management, and specifically to a lithium battery integrated management system based on Internet of Things and blockchain technology. Background Art

[0002] As the world accelerates its transition to low-carbon and sustainable energy, lithium batteries, as core energy storage components, are experiencing explosive growth in applications such as electric vehicles, smart grid energy storage, distributed energy systems, and various portable electronic devices. The performance, safety, and service life of lithium batteries are directly related to the stable operation and performance of these application systems.

[0003] At present, a variety of technical means are used for the management of lithium batteries:

[0004] Conventional data collection and monitoring: Traditional lithium battery management focuses on obtaining basic parameters, such as relying on simple voltage and current sensors with basic microcontrollers to regularly collect the voltage and current values ​​of the battery during charging and discharging, and then using preset fixed thresholds to determine whether the battery is in an overcharge or over-discharge state. In some simple application scenarios, the voltage is roughly monitored by the battery's own simple protection circuit, and the circuit is immediately cut off once it exceeds the limit to ensure basic safety. This type of primary data collection and monitoring mode can still meet the most basic safety control requirements in early small-scale, low-requirement lithium battery application scenarios.

[0005] Wired data transmission: In the data transmission link, in scenarios such as electric vehicles, the battery management system usually uses the CAN (Controller Area Network) bus to implement wired connection and transmit the collected data to the vehicle's main control unit. The CAN bus, with its high reliability and strong real-time performance, is suitable for short-distance data interaction in complex electromagnetic environments inside the vehicle, ensuring that the battery status can be mastered in real time during the vehicle's driving process. In the case of fixed energy storage facilities, RS485 serial communication is often used to aggregate data from multiple battery packs and transmit it to the monitoring center. This method is low-cost and convenient for large-scale networking deployment.

[0006] Basic battery status assessment: In the early days, when evaluating the battery status, most people relied on empirical formulas and simple mathematical models. For example, based on the roughly linear relationship between the battery open circuit voltage and the remaining power, the battery's SOC (State of Charge) is estimated by measuring the open circuit voltage; the ampere-hour integration method is used to integrate the charge and discharge currents in combination with the current sensor data to calculate the SOC, but this method has cumulative errors and requires regular calibration. As for the battery health status (SOH, State of Health) assessment, it is basically based on rough indicators such as the battery's service life and the number of charge and discharge times, and lacks precise quantitative analysis methods.

[0007] Although the existing lithium battery management technology has played a role at a certain stage, with the deepening of application and the upgrading of demand, many problems have become increasingly prominent:

[0008] First, data collection is partial and inaccurate. Traditional sensor configurations make it difficult to capture all aspects of lithium battery operation information. Parameters that have a significant impact on battery performance and life, such as battery internal resistance, temperature and humidity of the environment, and air pressure, are often omitted, making it difficult to accurately grasp the true state of the battery. At the same time, simple data collection methods are susceptible to electromagnetic interference, sensor aging and other factors, resulting in errors in the collected data, which cannot lay a solid foundation for subsequent analysis and evaluation, and thus interfere with the accuracy of battery management decisions.

[0009] Second, data transmission lacks flexibility and security. Although wired transmission is reliable in specific scenarios, the wiring is cumbersome and costly, which limits the scalability and deployment convenience of the battery system. Especially in scenarios such as distributed energy storage and mobile devices, it is difficult to meet the needs of rapid networking and flexible changes. Although wireless transmission has made progress, short-range communication technologies such as Bluetooth and Wi-Fi are unstable and easily cracked when facing complex environments and long-distance transmission. The integrity and confidentiality of data transmission cannot be guaranteed, and there is a risk of data leakage, which endangers the security of the battery system and even the entire application system.

[0010] Third, the battery status assessment methods are crude and lack precision and foresight. The assessment methods based on empirical formulas and simple models are difficult to match the complex electrochemical characteristics and diverse operating conditions of lithium batteries. The cumulative error of the ampere-hour integration method and the requirements of the open circuit voltage method for the static state of the battery make the SOC estimation accuracy low; the rough indicators of SOH assessment cannot provide insight into the potential failure risks of the battery in advance, and often wait until the battery performance declines significantly or the failure appears before taking action, which reduces the battery life and increases the operation and maintenance costs.

[0011] Therefore, a comprehensive lithium battery management system based on the Internet of Things and blockchain technology is needed to solve the above problems. Summary of the invention

[0012] Technical issues solved

[0013] In view of the shortcomings of the prior art, the present invention provides a lithium battery comprehensive management system based on the Internet of Things and blockchain technology, which solves the problems in the above background technology.

[0014] Technical Solution

[0015] To achieve the above objectives, the present invention is implemented through the following technical solutions: a lithium battery comprehensive management system based on the Internet of Things and blockchain technology, including a data acquisition module, a data transmission module, an edge-cloud collaborative processing module, an intelligent contract management module, a data security protection module, a cross-domain fusion module, a fault prediction module, an optimization scheduling module and a knowledge update module;

[0016] The data acquisition module collects the voltage, current, temperature, internal resistance and charge and discharge cycle times of the lithium battery in real time, and transmits the data to the edge node with the help of low-power wide area network technology. The data transmission module is responsible for uploading the preliminary data processed by the edge node to the cloud platform through the communication protocol to ensure the integrity and timeliness of data transmission;

[0017] The edge-cloud collaborative processing module integrates the adaptive Kalman filter algorithm and the deep neural network algorithm to perform preliminary noise reduction and feature extraction on the real-time data at the edge. The cloud uses the deep neural network for further analysis and processing to accurately assess the health status of the lithium battery. The smart contract management module combines hash locking and multi-signature technology to ensure the fairness of contract execution and financial security in the lithium battery leasing and second-hand trading business processes.

[0018] The data security protection module uses an algorithm that combines homomorphic encryption and zero-knowledge proof to encrypt and store sensitive data of lithium batteries, and in data sharing and verification scenarios, it protects privacy while proving the authenticity of data;

[0019] The cross-domain fusion module adopts distributed collaborative optimization algorithm and model prediction control algorithm to realize energy interaction and collaborative operation between lithium batteries and smart grids and intelligent transportation systems, and improve energy utilization efficiency. The fault prediction module uses gray prediction model and support vector machine regression algorithm to predict the type and time node of lithium battery faults in advance based on historical data, providing a basis for preventive maintenance.

[0020] The optimization scheduling module is based on multi-agent reinforcement learning and dynamic programming algorithms. It reasonably arranges the charging and discharging strategies of lithium batteries according to battery status, electricity demand and electricity price fluctuations to maximize economic benefits. The knowledge updating module has self-learning capabilities and automatically updates new data and model parameters.

[0021] Preferably, the data acquisition module simultaneously collects data such as the temperature, humidity and air pressure environment of the lithium battery environment, and transmits them synchronously with the core parameters to provide a data basis for multi-dimensional analysis.

[0022] Preferably, when processing data in the cloud, the edge-cloud collaborative processing module adopts a distributed computing framework to decompose large-scale data analysis tasks into multiple computing nodes for parallel processing, thereby improving processing efficiency.

[0023] Preferably, the smart contract management module has a built-in smart audit sub-module, which uses the traceability of blockchain to regularly audit the execution of contracts to ensure compliance operations.

[0024] Preferably, the data security protection module is provided with an access rights management submodule, which allocates corresponding data access rights to users and organizations based on role-based access control policies to prevent data leakage.

[0025] Preferably, when the cross-domain fusion module is integrated with the intelligent transportation system, it obtains the vehicle driving status information through the Internet of Vehicles technology, adjusts the energy output strategy of the lithium battery in real time, and ensures driving safety and energy-saving driving.

[0026] Preferably, the fault prediction report generated by the fault prediction module is displayed to the operation and maintenance personnel through a visual interface and is simultaneously pushed to the management system to facilitate timely arrangement of maintenance work.

[0027] Preferably, when formulating the charging and discharging strategy, the optimization scheduling module takes into account the aging characteristics of the battery, introduces a battery life loss model, and extends the battery life.

[0028] Preferably, after updating the model parameters, the knowledge updating module automatically performs model verification and uses test data to evaluate the performance of the updated model to ensure the effectiveness of the model.

[0029] Beneficial Effects

[0030] The present invention provides a lithium battery integrated management system based on the Internet of Things and blockchain technology. It has the following beneficial effects:

[0031] 1. The present invention overcomes the limitations of traditional sensor configurations by equipping it with a variety of high-precision sensors, such as voltage, current, temperature, internal resistance, ambient temperature and humidity, and air pressure sensors, and is able to capture lithium battery operating information in all directions. It can not only monitor the basic electrical parameters of the battery in real time, but also take environmental factors into consideration, providing a rich and accurate data basis for accurately grasping the true state of the battery, effectively avoiding management decision-making errors caused by one-sided data, and greatly improving the reliability and scientificity of battery management. In terms of data transmission, it integrates low-power wide area network technology, safe and reliable communication protocols, and signal conditioning circuits, combining flexibility and security. On the one hand, it gets rid of the cumbersome dilemma of wired transmission wiring, meets the needs of rapid networking and flexible changes in scenarios such as distributed energy storage and mobile devices, and improves the scalability of the system; on the other hand, relying on encryption technology and firewall protection, it effectively resists risks such as signal interference and hacker attacks.

[0032] 2. The present invention integrates adaptive Kalman filtering and deep neural network algorithms for health status assessment, and combines gray prediction model with support vector machine regression algorithm for fault prediction, saying goodbye to extensive assessment based on empirical formulas and simple models. These algorithms fully consider the complex electrochemical characteristics and diversified working conditions of lithium batteries, and effectively overcome the cumulative error of the ampere-hour integration method and the limitations of the open circuit voltage method.

[0033] 3. The present invention realizes efficient collaboration among system modules through innovative architectures such as edge-cloud collaborative processing and cross-domain integration. Edge-cloud collaboration enables the reasonable allocation of data processing tasks between the edge and the cloud, which can not only quickly respond to real-time needs at the edge, but also tap into the value of data with the help of the powerful computing power of the cloud; cross-domain integration promotes the energy interaction between lithium batteries and smart grids and intelligent transportation systems, improves energy utilization efficiency, and optimizes system performance from multiple dimensions.

[0034] 4. The smart contract management module in the present invention combines hash locking and multi-signature technology to provide a fair, transparent and tamper-proof execution environment for lithium battery leasing, second-hand transactions and other businesses, effectively resolving transaction disputes and enhancing market trust. At the same time, the data security protection module uses homomorphic encryption and zero-knowledge proof technology to meet regulatory requirements and effectively protect the privacy of enterprises and users in data sharing and verification scenarios, ensuring the security of sensitive data and creating a good ecological environment for the healthy development of the lithium battery industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a system flow chart of the present invention;

[0036] Figure 2 It is a hardware composition diagram of the system module of the present invention. DETAILED DESCRIPTION

[0037] 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:

[0039] like Figure 1-2 As shown in the figure, the lithium battery comprehensive management system based on the Internet of Things and blockchain technology includes data acquisition module, data transmission module, edge-cloud collaborative processing module, smart contract management module, data security protection module, cross-domain fusion module, fault prediction module, optimization scheduling module and knowledge update module;

[0040] The data acquisition module collects the voltage, current, temperature, internal resistance and number of charge and discharge cycles of the lithium battery in real time, and transmits the data to the edge node with the help of low-power wide area network technology. The data transmission module is responsible for uploading the preliminary data processed by the edge node to the cloud platform through the communication protocol to ensure the integrity and timeliness of data transmission; the edge-cloud collaborative processing module integrates the adaptive Kalman filter algorithm and the deep neural network algorithm to perform preliminary noise reduction and feature extraction on the real-time data at the edge. The cloud uses deep neural networks for further analysis and processing to accurately evaluate the health status of the lithium battery. The smart contract management module combines hash locking and multi-signature technology to ensure the fairness of contract execution and financial security in the lithium battery leasing and second-hand trading business processes; the data security protection module uses homomorphic encryption and zero-knowledge proof The algorithm combines the use of the smart grid and the intelligent transportation system to encrypt and store the sensitive data of the lithium battery. In the data sharing and verification scenario, it protects privacy while proving the authenticity of the data. The cross-domain fusion module adopts the distributed collaborative optimization algorithm and the model prediction control algorithm to realize the energy interaction and collaborative operation between the lithium battery and the smart grid and the intelligent transportation system, and improve the energy utilization efficiency. The fault prediction module uses the grey prediction model and the support vector machine regression algorithm to predict the type and time node of the lithium battery fault in advance according to the historical data, and provide a basis for preventive maintenance. The optimization scheduling module is based on multi-agent reinforcement learning and dynamic programming algorithms. According to the battery status, power demand and electricity price fluctuations, the charging and discharging strategy of the lithium battery is reasonably arranged to maximize the economic benefits. The knowledge update module has the ability of self-learning and automatically updates the new data and model parameters.

[0041] The data acquisition module simultaneously collects data such as the temperature, humidity and air pressure of the lithium battery's environment, and transmits them synchronously with the core parameters to provide a data basis for multi-dimensional analysis; when the edge-cloud collaborative processing module processes data in the cloud, it adopts a distributed computing framework to decompose large-scale data analysis tasks into multiple computing nodes for parallel processing, thereby improving processing efficiency; the smart contract management module has a built-in intelligent audit sub-module, which uses the traceability of blockchain to regularly audit the execution of contracts to ensure compliance operations; the data security protection module sets up an access rights management sub-module, which allocates corresponding data access rights to users and institutions based on role-based access control policies to prevent data leakage.

[0042] When the cross-domain fusion module is integrated with the intelligent transportation system, it obtains the vehicle driving status information through the Internet of Vehicles technology, adjusts the energy output strategy of the lithium battery in real time, and ensures driving safety and energy-saving driving; the fault prediction report generated by the fault prediction module is displayed to the operation and maintenance personnel through a visual interface, and is pushed to the management system at the same time to facilitate timely arrangement of maintenance work; when formulating the charging and discharging strategy, the optimization scheduling module considers the aging characteristics of the battery, introduces a battery life loss model, and extends the battery life; after updating the model parameters, the knowledge update module automatically performs model verification, and uses test data to evaluate the performance of the updated model to ensure the effectiveness of the model. Specific embodiment one:

[0044] like Figure 1-2 As shown, the workflow of the entire system is as follows:

[0045] Data collection starts:

[0046] The data acquisition module starts working, and its built-in high-precision sensors continuously and in real time collect the core parameters of the lithium battery, such as voltage, current, temperature, internal resistance, and number of charge and discharge cycles. At the same time, the environmental monitoring sensors connected to it synchronously collect data such as the temperature, humidity, and air pressure of the environment where the lithium battery is located. These sensors have been calibrated and optimized, with high sensitivity and low error characteristics to ensure the accuracy of the collected data.

[0047] The collected data is initially packaged and organized in the data acquisition module. According to the preset data format and encoding rules, various parameters under the same timestamp are integrated into a data unit to prepare for subsequent transmission.

[0048] Edge node data preprocessing:

[0049] Through low-power wide area network technologies (such as NB-IoT and LoRa), the data acquisition module quickly transmits the sorted data units to the edge nodes that are close to the data source. After the edge node receives the data, the edge-cloud collaborative processing module immediately starts the task at the edge.

[0050] First, the adaptive Kalman filter algorithm is used to perform preliminary noise reduction on real-time data. The algorithm can adjust the filter parameters in real time according to the dynamic change characteristics of the data, effectively remove the noise introduced by factors such as sensor accuracy and electromagnetic interference, and restore the real fluctuation trend of the data.

[0051] Next, the shallow structure of the deep neural network algorithm is used to extract features from the denoised data. Based on the pre-trained model for lithium battery parameter characteristics, key features such as battery charge and discharge curve characteristics and temperature change slope are quickly identified, providing more valuable preliminary data for further analysis in the cloud.

[0052] In-depth analysis and health assessment in the cloud:

[0053] The data initially processed by the edge node is uploaded to the cloud platform by the data transmission module according to a secure and reliable communication protocol (such as HTTPS encryption protocol). After the cloud platform receives the data, the cloud part of the edge-cloud collaborative processing module begins to operate.

[0054] The cloud uses the deep and complex structure of the deep neural network algorithm, combined with the characteristic data transmitted from the edge and the massive historical data stored in the cloud, to accurately evaluate the health status of the lithium battery. Through multi-dimensional analysis of the battery capacity attenuation law, internal resistance change trend, abnormal temperature fluctuations and the relationship between battery failure, it outputs detailed health assessment indicators such as the percentage of remaining available battery capacity, health status score (0-100 points), and estimated remaining life.

[0055] Smart contract business guarantee:

[0056] When it comes to lithium battery leasing and second-hand trading business, the smart contract management module plays a key role. When business needs are triggered (such as users initiating leasing applications or sellers publishing second-hand battery sales information), smart contracts are created by combining hash locking and multi-signature technology.

[0057] Hash locking ensures that key transaction information (such as battery specifications, rental duration and price) cannot be tampered with during the execution of the contract. Multi-signature requires multiple parties involved in the transaction (such as lessors, lessees, platform regulators, etc.) to sign and confirm in accordance with the agreed order and rules. Only when all legal signatures are collected, the contract funds will flow according to the predetermined process to ensure the fairness of contract execution and the security of funds.

[0058] At the same time, the built-in intelligent audit submodule uses the traceability of blockchain to conduct regular and comprehensive audits on contract execution. From contract creation, capital flow, battery delivery to usage feedback, the authenticity and compliance of data can be traced. If anomalies are found, an immediate warning will be issued and the investigation process will be initiated.

[0059] Data security protection throughout the process:

[0060] The data security protection module is involved at the beginning of data collection. Using an algorithm that combines homomorphic encryption with zero-knowledge proof, sensitive data of lithium batteries is encrypted and stored. At the data collection end, the data collected by the sensor is encrypted as soon as it is generated to ensure the confidentiality of the data during transmission and storage.

[0061] When there is a need for data sharing verification (such as random inspections by regulatory authorities and cooperative research by scientific research institutions), zero-knowledge proof technology is used to prove the authenticity of the data to the verification party without leaking the original data content. At the same time, the access rights management submodule allocates corresponding data access rights to different users (such as ordinary operation and maintenance personnel, senior engineers, corporate management and external regulatory agencies) and institutions based on a sophisticated role-based access control strategy. Data leakage is prevented through multiple levels such as identity authentication and permission review.

[0062] Cross-domain fusion energy interaction:

[0063] The cross-domain fusion module realizes the energy interaction and coordinated operation between lithium batteries and smart grids and intelligent transportation systems. Taking the integration with intelligent transportation systems as an example, through the Internet of Vehicles technology (such as V2X communication), real-time vehicle driving status information (including vehicle speed, acceleration, driving route, and traffic congestion status data) is obtained.

[0064] Based on this information, the distributed collaborative optimization algorithm and the model predictive control algorithm are combined to adjust the energy output strategy of the lithium battery in real time. In the acceleration and overtaking scenario, the battery power is reasonably allocated to ensure the vehicle's power needs; when driving slowly in congestion, the battery output power is reduced to reduce unnecessary energy loss and achieve energy-saving driving. On the other hand, when the vehicle battery is fully charged and the grid load is at a peak, the excess power of the vehicle battery is fed back to the grid through the bidirectional charging and discharging device, which improves energy utilization efficiency and helps balance the supply and demand of the smart grid.

[0065] Fault prediction and early prevention:

[0066] The fault prediction module uses the grey prediction model and support vector machine regression algorithm to predict the possible fault types and time nodes of lithium batteries based on the massive historical data stored in the cloud. The grey prediction model is good at processing small samples and poor information data, mining the potential trends of the data, and preliminarily predicting the approximate time range of the fault.

[0067] On this basis, the support vector machine regression algorithm combines the current real-time parameters and environmental data of the battery to further refine the fault type (such as short circuit fault, thermal runaway fault and capacity drop fault). The generated fault prediction report is displayed to the operation and maintenance personnel through a visual interface (such as the Web dashboard and mobile APP push notification), and is also pushed to relevant management systems (such as the enterprise equipment management system and the power grid dispatching system) to facilitate the timely arrangement of maintenance work.

[0068] Optimize scheduling and improve efficiency:

[0069] The optimization scheduling module is based on multi-agent reinforcement learning and dynamic programming algorithms to continuously monitor battery status, power demand and electricity price fluctuations. The multi-agent reinforcement learning algorithm simulates the interactive decision-making process of multiple agents (such as battery management units, power equipment and grid access nodes) in different environments, and learns the optimal charging and discharging strategy through continuous trial and error and reward feedback mechanism.

[0070] On this basis, the dynamic programming algorithm takes into account the aging characteristics of the battery, introduces a battery life loss model, and weighs the current charging and discharging benefits and battery life loss costs from the perspective of long-term economic benefits. During the low electricity price period, the battery charging is reasonably arranged to store more electricity; during the peak electricity consumption and high electricity prices, the battery discharge is controlled to meet the electricity demand while earning the price difference to maximize economic benefits.

[0071] Knowledge update continues to evolve:

[0072] The knowledge update module has self-learning capabilities, and automatically starts the update process as new data continues to flow in during system operation. First, according to the preset data screening rules, representative and high-quality new data are selected and integrated with historical data in the cloud.

[0073] Then, the fused data is used to retrain and update the parameters of the models involved in each module (such as deep neural network models, fault prediction models, and optimization scheduling models). After the update is completed, the model is automatically verified, and the performance of the updated model is evaluated using independent test data. By comparing the evaluation indicators before and after the update (such as prediction accuracy, error rate, and economic benefit improvement), the model effectiveness is ensured, so that the entire system continues to adapt to the dynamic changes of lithium batteries. Specific embodiment three:

[0075] like Figure 1-2 As shown, the key algorithm mentioned in Example 1 is analyzed in detail below, including its core mathematical formula and explanation:

[0076] 1. Adaptive Kalman Filter Algorithm

[0077] Core mathematical formula:

[0078] State prediction equation:

[0079]

[0080] in, is the prior state estimate at time k, F k It is the state transfer matrix, which describes the state change law of the system from time k-1 to time k; is the posterior state estimate at time k-1; B k is the control input matrix, uk-1 is the control input vector, which can be ignored if there is no external control input.

[0081] Covariance prediction equation:

[0082] here, is the prior estimate covariance matrix, which is used to measure the uncertainty of state estimation; P k-1 is the posterior estimated covariance matrix of the previous moment; Q k is the process noise covariance matrix, which represents the uncertainty in the system modeling process, such as sensor noise and unmodeled dynamic characteristics.

[0083] Measurement update equation:

[0084]

[0085] Among them, K k is the Kalman gain, which determines the weight distribution of the measured value and the predicted value in the final state estimation; H k is the measurement matrix, which maps the state space to the measurement space; z k is the measured value at time k; R k is the measurement noise covariance matrix, which represents the uncertainty in the measurement process; is the posterior state estimate at time k, that is, the optimal estimate after measurement update; P k is the posterior estimated covariance matrix.

[0086] In lithium battery data processing, the adaptive Kalman filter algorithm first estimates the state of the previous moment. And the state transition model of the system (through F k Predict the current state Considering the process noise (Q k ) on the uncertainty of the forecast, and obtain the prior estimated covariance When the sensor measurement value (z k ) and then, through the measurement matrix (H k ) maps the predicted state to the measurement space and compares it with the actual measurement value. k ) Dynamically adjust the weights of the measured and predicted values ​​so that the final state estimate The optimal balance between the predicted value and the measured value is achieved, noise interference is effectively reduced, and the real fluctuation trend of the data is restored. In addition, the algorithm adaptively adjusts the process noise covariance (Q k ) and the measurement noise covariance (R k ), in order to adapt to the dynamic changes in the operation of lithium batteries, such as changes in data characteristics caused by factors such as battery aging and changes in operating conditions.

[0087] 2. Deep Neural Network Algorithm

[0088] Forward propagation: For a multi-layer perceptron (taking three layers as an example), the input layer neurons receive data x, and the output calculation formula of the hidden layer neurons is:

[0089]

[0090] Where j = 1, 2, ..., m, σ is the activation function (such as ReLU function: σ(x) = max(0, x)), is the weight from the input layer to the hidden layer, is the bias of the hidden layer, n is the number of neurons in the input layer, and m is the number of neurons in the hidden layer.

[0091] The output calculation formula of the output layer neurons is:

[0092]

[0093] where k = 1, 2, ..., p, is the weight from the hidden layer to the output layer, is the bias of the output layer, and p is the number of neurons in the output layer.

[0094] Back propagation: Update weights based on the gradient descent method, with the mean square error loss function as For example (where is the true label),

[0095] Calculate the output layer weight gradient:

[0096]

[0097] The weight gradient from the input layer to the hidden layer can also be calculated, and then the gradient descent formula is:

[0098]

[0099] Update the weights, where η is the learning rate.

[0100] 3. Homomorphic encryption and zero-knowledge proof technology

[0101] The core mathematical formula of homomorphic encryption is:

[0102] Taking a simple additive homomorphic encryption algorithm Paillier encryption system as an example, let the plaintext be m1, m2, the public key be (n, g), and the private key be λ.

[0103] Cryptographic operations: Among them, r1 and r2 are random numbers.

[0104] Homomorphic addition: c1·c2modn 2 =E(m1+m2), that is, multiplication operation can be directly performed on the ciphertext to achieve additive homomorphism of the plaintext.

[0105] In the scenario of protecting sensitive lithium battery data, homomorphic encryption allows specific operations to be performed on data in an encrypted state. For example, the cloud needs to count the total number of charge and discharge times of a batch of lithium batteries. The data of each battery is encrypted locally (such as using Paillier encryption) and then uploaded. The cloud can obtain the ciphertext result of the total number of charge and discharge times through ciphertext multiplication operations (corresponding to plaintext addition) without decryption, which not only protects data privacy, but also meets specific data processing requirements.

[0106] 4. Distributed Collaborative Optimization Algorithm

[0107] Distributed collaborative optimization based on the consensus algorithm. Assume that there are N nodes, and each node i holds a local variable x. i , the goal is to minimize the global objective function

[0108]

[0109] where f i (x) is the local objective function of node i. In each iteration, nodes exchange information through communication. The update rule is as follows:

[0110]

[0111] Where k is the number of iterations, w ij (k) is the influence weight of node j on node i at time k (satisfying α is the step size, is the gradient of the local objective function of node i at time k.

[0112] When lithium batteries work together with smart grids and intelligent transportation systems, different nodes (such as distributed energy generation units in the power grid, vehicle battery management units in the transportation system, etc.) have different optimization goals (such as minimizing power generation costs, minimizing vehicle energy consumption, etc.). However, they need to work together to improve the global energy efficiency. Through the distributed collaborative optimization algorithm, nodes communicate with each other and use the weight matrix (w ij (k)) Fusion of other node information and its own gradient information ) Updates its own variable (x i (k+1)), gradually approaching the global optimal solution and realizing the reasonable allocation and coordinated operation of energy.

[0113] 5. Model Predictive Control Algorithm

[0114] Prediction model: A linear time-invariant model is used here, and the state space expression is:

[0115] x k+1 =Ax k +Bu k +w k ,y k =Cx k +Du k +v k

[0116] Among them, x k is the state vector at time k (such as the battery power, temperature, etc.), u k is the control input vector (such as charge and discharge current), y k is the output vector (such as battery voltage, power and other measurable values), A, B, C, D are system matrices, and w k 、v k are process noise and measurement noise respectively.

[0117] Optimization goal: Minimize the objective function within the limited prediction time domain N:

[0118]

[0119] Among them, r k is the reference trajectory (such as the desired battery output power trajectory), Q, R, and P are weight matrices that weigh the importance of output tracking error, control input energy, and terminal state error, respectively.

[0120] Solution: Solve the above optimization problem through dynamic programming and other methods to obtain the optimal control sequence And the first control input Applied to the system.

[0121] In the lithium battery energy output strategy adjustment scenario, such as vehicle driving, the model predictive control firstly calculates the current state (x k ) and control input (u k ), using the prediction model to estimate the system state in the future (x k+1 ) and output (y k ). Combined with the reference trajectory (r k ), by minimizing the objective function (J), weighing factors such as tracking error and control energy consumption, and using dynamic programming to solve the optimal control sequence. For example, when accelerating to overtake, the power reference trajectory is set according to the vehicle power demand, and the algorithm solves the appropriate charge and discharge current Ensure power supply while taking into account battery life and energy consumption to achieve energy-saving driving. Specific embodiment four:

[0123] like Figure 1-2As shown, the following is a detailed hardware composition and hardware description of each module in Example 1:

[0124] 1. Data acquisition module

[0125] Sensor Group:

[0126] Voltage sensor: A high-precision voltage divider resistor voltage sensor is used, such as the voltage measurement circuit composed of the LM358 operational amplifier of Texas Instruments (TI) and precision resistors. It can accurately measure the terminal voltage of the lithium battery. The measurement range is usually 0-5V, and the accuracy can reach ±0.01V. It can meet the voltage monitoring needs of lithium batteries of different specifications and convert the voltage signal into an analog electrical signal that can be processed by the microcontroller.

[0127] Current sensor: The current sensor based on the Hall effect principle can measure from -30A to +30A with an accuracy of about ±1.5%. It can detect the charging and discharging current of lithium batteries in real time, sense the current through changes in the magnetic field, and output a voltage signal proportional to the current, so as to accurately monitor the charging and discharging status of the battery.

[0128] Temperature sensor: Choose a negative temperature coefficient (NTC) thermistor sensor, such as Vishay's NTCLE100E3103JB0 model. Its resistance value changes with temperature. It has the characteristics of high sensitivity and fast response. The measurement temperature range is generally -40℃ to +125℃, with an accuracy of up to ±0.5℃. It is used to monitor the temperature of lithium batteries during operation to prevent safety problems caused by overheating.

[0129] Internal resistance sensor: A professional internal resistance test chip that uses the AC injection measurement principle, such as Analog Devices' LTC4020. It injects a small AC signal into the battery and calculates the internal resistance by measuring the voltage response and current response at both ends of the battery. The measurement accuracy can reach milliohms, which can promptly reflect the internal aging, short circuit and other conditions of the battery, providing key data for battery health assessment.

[0130] Charge and discharge cycle counter: It is essentially a counter chip with power-off memory function, such as Maxim Integrated's DS1621. Each time the battery completes a complete charge and discharge process, the counter automatically accumulates 1 to record the usage history of the lithium battery and assist in determining the remaining battery life.

[0131] Microcontroller (MCU): Choose low-power, high-performance ARMCortex-M series MCU, such as the STM32L4 series, which integrates a wealth of peripheral interfaces, including multiple ADC (analog-to-digital converter) channels for connecting sensors to collect analog signals, SPI (serial peripheral interface), I2C (inter-integrated circuit) and other communication interfaces for communicating with other modules. It has powerful data processing capabilities and can pre-process and package the collected sensor data in a short time for subsequent transmission.

[0132] Wireless transmission module:

[0133] Bluetooth Low Energy (BLE) Module: Like Nordic Semiconductor's nRF52832 chip, it supports the Bluetooth 5.0 standard, has the characteristics of low power consumption and moderate transmission rate (up to 2Mbps), and is suitable for close-range and small data volume scenarios, such as transmitting collected data to nearby edge nodes or mobile terminal devices for preliminary viewing.

[0134] Low-power wide area network (LPWAN) module: Taking LoRaWAN as an example, it uses Semtech's SX1276 / 78 chip, works in the Sub-GHz frequency band, has a long communication distance (visible distance can reach more than ten kilometers), extremely low power consumption, can penetrate obstacles such as buildings, and reliably transmit data to edge nodes, meeting the data collection and transmission needs of large-scale lithium battery packs in complex environments.

[0135] 2. Data transmission module

[0136] Communication interface chip:

[0137] Ethernet controller chip: such as Realtek's RTL8211E, which supports high-speed Ethernet connection (10 / 100 / 1000Mbps). When the system needs to access the wired network to upload data to the cloud platform, the chip is responsible for processing the data frame transmission and reception at the bottom of the network protocol stack, communicating with the host computer or network switch to ensure high-speed and stable data transmission.

[0138] 4G / 5G module: Choose EC20 / EC50 series 4G modules or 5G modules such as RM500Q-GL from Quectel. They have built-in complete mobile communication protocol stacks, can be directly inserted into SIM cards, use operator networks to achieve long-distance, high-speed data transmission, and quickly upload data processed by edge nodes to the cloud, adapting to application scenarios under different network coverage conditions.

[0139] Signal conditioning circuit: It is composed of operational amplifiers, filters, etc., and is used to amplify and filter the signals during data transmission. For example, TI's OPA2134 high-performance operational amplifier is used to amplify the weak current signal from the data acquisition module to meet the input level requirements of the communication interface chip; at the same time, an RC filter circuit is used to remove high-frequency noise in the signal to ensure the accuracy of data transmission.

[0140] 3. Edge-cloud collaborative processing module (edge ​​end)

[0141] Edge computing devices:

[0142] Industrial-grade gateway: You can choose Huawei's AR502EGRc-Lc industrial gateway, which has powerful edge computing capabilities, is equipped with a multi-core processor (such as the ARMCortex-A7 series), and has built-in multiple communication interfaces (Ethernet, serial port, USB, Wi-Fi, and Bluetooth). On the one hand, it can receive data from the data acquisition module, and on the other hand, it can communicate with the cloud platform through wired or wireless networks. At the same time, it supports containerized deployment of applications, which facilitates the operation of software programs based on algorithms such as adaptive Kalman filtering and shallow feature extraction of deep neural networks, and performs preliminary noise reduction and feature extraction on real-time data.

[0143] Field Programmable Gate Array (FPGA): For some scenarios with extremely high real-time requirements and large algorithm calculations, Xilinx's Spartan-7 series FPGA can be used. FPGA can dynamically program hardware circuits according to specific algorithm requirements, has strong parallel data processing capabilities, can quickly complete matrix operations such as in the adaptive Kalman filter algorithm, accelerate data processing speed, and ensure that preliminary processing results are given in a timely manner at the edge.

[0144] 4. Edge-Cloud Collaborative Processing Module (Cloud)

[0145] Cloud server: Use general-purpose cloud server instances from mainstream cloud service providers (such as Alibaba Cloud, Tencent Cloud, Amazon AWS, etc.), configure multi-core CPUs (such as Intel Xeon series), large-capacity memory (such as 32GB or more), and high-speed solid-state drives (such as 1TB or more), run software services based on deep neural network deep structure analysis, distributed computing frameworks, etc., use the powerful computing resources of the cloud platform to conduct in-depth analysis and processing of data transmitted from the edge, and accurately evaluate the health status of lithium batteries.

[0146] Distributed storage devices: Distributed file systems based on cloud platforms (such as Ceph, GlusterFS, etc.) or object storage services (such as AWSS3, Alibaba Cloud OSS, etc.) are used to store massive amounts of lithium battery historical data, model parameters, transaction records and other information. They have high reliability (data redundancy technology is used to ensure that data is not lost) and high scalability (storage capacity can be easily increased as the amount of data grows), providing data support for data analysis, smart contract execution, fault prediction and other functions.

[0147] 5. Smart Contract Management Module

[0148] Blockchain node server:

[0149] Ethereum private chain node server: Use high-performance servers to build Ethereum private chain nodes, such as Dell PowerEdge R740, equipped with dual-core Intel Xeon scalable processors and large-capacity memory (more than 64GB), run Ethereum client software (such as Geth, Parity, etc.), and be responsible for the creation, deployment, execution and maintenance of smart contracts, ensuring the fairness, immutability and financial security of contract execution, and using the distributed ledger characteristics of blockchain to record key information in business processes such as lithium battery leasing and second-hand transactions.

[0150] Blockchain hardware wallets: Hardware wallets such as Ledger Nano S or Trezor are used to store users' private keys. Compared with software wallets, they have higher security and prevent private keys from being stolen by hackers. In the fund transaction process involved in smart contracts, digital signature operations are performed through hardware wallets to ensure that only the legal owner can authorize the transfer of funds.

[0151] 6. Data security protection module

[0152] Encryption chip:

[0153] National SM4 encryption chip: Ruxin Dajiean's XS100 series chips use the domestically produced independently controllable SM4 encryption algorithm to encrypt and store sensitive data of lithium batteries. Whether at the data collection end, transmission process, or storage at the edge node or cloud, it can ensure the confidentiality of data and prevent data leakage, meeting the high domestic standards for information security.

[0154] Homomorphic encryption dedicated chips: Some scientific research institutions and enterprises are developing homomorphic encryption dedicated chips, such as experimental chips based on the Paillier algorithm, which can accelerate the homomorphic encryption calculation process. When encrypted data needs to be processed in the cloud (such as statistical analysis), it can improve computing efficiency and realize the effective use of data in an encrypted state while protecting privacy.

[0155] Firewall equipment: Choose an enterprise-level firewall, such as the Cisco ASA series, and deploy it at the connection between the cloud platform and the external network and at the access point between the edge node and the external network. By setting access control policies, based on conditions such as source IP address, destination IP address, port number, and protocol type, block illegal external network access, prevent security threats such as hacker attacks and malware intrusions, and protect the data security of the entire system.

[0156] 7. Cross-domain integration module

[0157] Internet of Vehicles communication module (when integrated with intelligent transportation system):

[0158] On-board unit (OBU): Uses dedicated Internet of Vehicles communication chips, such as NXP's SAF5000 series, supports DSRC (dedicated short-range communication) or C-V2X (cellular vehicle-to-everything) technology, enables information exchange between vehicles and roadside infrastructure (such as charging piles, traffic lights), and other vehicles, and obtains vehicle driving status information (speed, acceleration, driving route, traffic congestion conditions, etc.) in real time, providing a basis for adjusting the lithium battery energy output strategy.

[0159] V2X Roadside Unit (RSU): RSU products provided by Huawei, Datang and other manufacturers are usually installed on roadside traffic facilities and have strong wireless communication capabilities. On the one hand, they communicate with the on-board unit to receive vehicle information; on the other hand, they connect to the traffic management system to integrate vehicle and road information to assist in traffic decision-making. At the same time, they also provide comprehensive traffic scenario data for cross-domain fusion modules, promoting the coordinated operation of lithium batteries and intelligent transportation systems.

[0160] 8. Fault prediction module

[0161] Data analysis server: Similar to the cloud server in the edge-cloud collaborative processing module (cloud), it is equipped with a high-performance CPU, large-capacity memory and high-speed storage, and runs data analysis software based on grey prediction models, support vector machine regression algorithms, etc. It uses the massive historical data stored in the cloud to predict the possible failure types and time nodes of lithium batteries in advance, providing a basis for preventive maintenance.

[0162] Data visualization equipment: Displays or large-screen splicing display systems connected to data analysis servers, such as Dell UltraSharp series displays or Barco splicing screens, use visualization software (such as Tableau, PowerBI, etc.) to display fault prediction reports to operation and maintenance personnel in the form of intuitive charts and graphs, allowing them to quickly understand the risk of battery failures and push them to relevant management systems to facilitate timely scheduling of maintenance work.

[0163] 9. Optimize the scheduling module

[0164] Smart meters: High-precision smart meters, such as Schneider Electric's ION7650 series, are used to measure electricity demand in real time, accurate to 0.1 kWh. They can not only measure basic parameters such as active power, reactive power, current, and voltage, but also have communication functions. They can transmit electricity consumption data in real time to the optimization scheduling module through communication protocols such as Modbus, providing accurate electricity consumption information for formulating charging and discharging strategies.

[0165] Electricity price monitoring equipment: Real-time information on electricity price fluctuations can be obtained through web crawler technology or dedicated software and hardware equipment connected to the electricity trading platform, such as Jinzhi Technology's electricity market transaction auxiliary decision-making system, which can parse electricity market price data, update electricity prices every 15 minutes, and provide real-time electricity price data to the optimization scheduling module, so as to reasonably arrange battery charging during the period of low electricity prices and maximize economic benefits.

[0166] 10. Knowledge Update Module

[0167] Data acquisition server: It is also a server with certain computing power and storage capacity. It is used to receive new data. Before the data enters the system, it selects representative and high-quality new data according to the preset data screening rules and integrates it with the historical data in the cloud to provide sufficient and high-quality data resources for subsequent model training.

[0168] Model training server: It is configured with high-performance hardware similar to cloud servers, runs deep learning frameworks (such as TensorFlow, PyTorch, etc.) and machine learning algorithm libraries (such as Scikit-learn, etc.), and uses the fused data to retrain and update the parameters of the models involved in each module (such as deep neural network models, fault prediction models, optimization scheduling models, etc.). After the update is completed, the model is automatically verified, and the performance of the updated model is evaluated using independent test data to ensure the effectiveness of the model, so that the entire system can continue to adapt to the dynamic changes of lithium batteries.

[0169] 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.

[0170] 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. A comprehensive lithium battery management system based on the Internet of Things and blockchain technology, characterized by : It includes data acquisition module, data transmission module, edge-cloud collaborative processing module, smart contract management module, data security protection module, cross-domain fusion module, fault prediction module, optimization scheduling module and knowledge update module; The data acquisition module collects the voltage, current, temperature, internal resistance and charge and discharge cycle times of the lithium battery in real time, and transmits the data to the edge node with the help of low-power wide area network technology. The data transmission module is responsible for uploading the preliminary data processed by the edge node to the cloud platform through the communication protocol; The edge-cloud collaborative processing module is divided into the edge and the cloud, integrating the adaptive Kalman filter algorithm and the deep neural network algorithm, performing preliminary noise reduction and feature extraction on the real-time data at the edge, and further analyzing and processing using the deep neural network on the cloud to accurately assess the health status of the lithium battery. The smart contract management module combines hash locking and multi-signature technology to ensure the fairness of contract execution and financial security in the lithium battery leasing and second-hand trading business processes; The data security protection module uses an algorithm that combines homomorphic encryption and zero-knowledge proof to encrypt and store sensitive data of lithium batteries, and in data sharing and verification scenarios, it protects privacy while proving the authenticity of data; The cross-domain fusion module adopts distributed collaborative optimization algorithm and model prediction control algorithm to realize energy interaction and collaborative operation between lithium batteries and smart grids and intelligent transportation systems, and improve energy utilization efficiency. The fault prediction module uses gray prediction model and support vector machine regression algorithm to predict the type and time node of lithium battery faults in advance based on historical data. The optimization scheduling module is based on multi-agent reinforcement learning and dynamic programming algorithms. It reasonably arranges the charging and discharging strategies of lithium batteries according to battery status, electricity demand and electricity price fluctuations to maximize economic benefits. The knowledge updating module has self-learning capabilities and automatically updates new data and model parameters.

2. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: The data acquisition module simultaneously collects data such as the temperature, humidity and air pressure environment of the lithium battery environment, and transmits them synchronously with parameters such as voltage, current, temperature, internal resistance and number of charge and discharge cycles.

3. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized in that: When processing data in the cloud, the edge-cloud collaborative processing module adopts a distributed computing framework to decompose large-scale data analysis tasks into multiple computing nodes for parallel processing, thereby improving processing efficiency.

4. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: The smart contract management module has a built-in smart audit sub-module, which uses the traceability of blockchain to regularly audit the execution of contracts to ensure compliance operations.

5. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: The data security protection module is provided with an access rights management submodule, which allocates corresponding data access rights to users and institutions based on role-based access control policies to prevent data leakage.

6. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: When the cross-domain fusion module is integrated with the intelligent transportation system, it obtains vehicle driving status information through the Internet of Vehicles technology and adjusts the energy output strategy of the lithium battery in real time.

7. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: The fault prediction report generated by the fault prediction module is displayed to the operation and maintenance personnel through a visual interface and is pushed to the management system at the same time.

8. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: When formulating the charging and discharging strategy, the optimization scheduling module takes into account the aging characteristics of the battery, introduces a battery life loss model, and prolongs the battery life.

9. The lithium battery integrated management system based on the Internet of Things and blockchain technology according to claim 1 is characterized by: After updating the model parameters, the knowledge updating module automatically performs model verification and uses test data to evaluate the performance of the updated model to ensure the effectiveness of the model.

Images