Vehicle-mounted power supply control system based on MCU
By adopting MCU and neural network models in the on-board power control system, dynamically allocating power supplies and using blockchain technology, the problems of inaccurate load demand prediction and poor integrity of fault data in the existing technology are solved, and more efficient and safe Vehicle power management.
Patent Information
- Application Number
- CN202411131974.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing vehicle power control system has low accuracy in predicting future load demands, and cannot respond to changes in demand of electronic systems in a timely manner. Manual intervention has increased, poor performance and fault data integrity of vehicle electronic systems, and high risk of single point failure.
A vehicle power control system based on MCU is designed, including power input module, converter, monitoring and management module, load management module, energy saving decision-making module, fault diagnosis module, remote communication module, energy integration module and record analysis module, through monitoring Battery status and vehicle electronic system load requirements, dynamic distribution of power, use neural network models to predict power demand, and combine blockchain technology to ensure data security and integrity.
It improves the accuracy of load demand forecasting, reduces manual intervention, enhances the performance and failure data integrity of the vehicle electronic system, reduces the risk of single point of failure, and improves the intelligence and safety of the system.
Smart Images

Figure CN119051256A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power management control, and in particular to an MCU-based vehicle power control system. Background Art
[0002] With the continuous development of modern automobiles, on-board power systems are becoming more and more complex and critical in vehicles. These systems are responsible for powering driving, infotainment, safety, communications, and various other electronic devices. In order to meet these diverse needs, the vehicle power system not only needs to provide sufficient power, but also requires intelligent control and management to ensure efficient, reliable and safe operation. In addition, with the popularity of electric vehicles, battery management and energy distribution have become more important challenges. MCU-based on-board power control system came into being, providing an intelligent solution for the vehicle power system.
[0003] The existing vehicle power control system has low accuracy in predicting future load demand and cannot respond to changes in electronic system demand in a timely manner, increasing the need for manual intervention. In addition, the performance of the vehicle electronic system and the integrity of fault data of the existing vehicle power control system are poor, and the information among the participants is asymmetric, resulting in a high risk of single point failure. Therefore, we propose an MCU-based vehicle power control system. Summary of the invention
[0004] The purpose of the present invention is to solve the defects in the prior art and to propose an MCU-based vehicle power control system.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions:
[0006] An MCU-based vehicle power control system includes a power input module, a converter, a monitoring management module, a load management module, an energy-saving decision module, a fault diagnosis module, a remote communication module, an energy integration module, a recording and analysis module, and a user platform;
[0007] The power input module is used to receive power input from the vehicle battery and the generator;
[0008] The converter is used to convert power from the battery and the generator into a constant voltage and current;
[0009] The monitoring and management module is used to monitor the battery status and intelligently control charging and discharging;
[0010] The load management module monitors the load requirements of each vehicle electronic system and dynamically allocates power;
[0011] The energy-saving decision module is used to monitor the energy efficiency of the vehicle and generate energy-saving decisions, while managing the start and stop of the auxiliary system;
[0012] The fault diagnosis module is used to detect power system faults to provide real-time alarms and remote diagnosis;
[0013] The remote communication module is used to communicate with the vehicle's interconnected systems and receive remote commands and updates;
[0014] The energy integration module is used to integrate renewable energy to provide additional power;
[0015] The recording and analysis module is used to record power system performance data, and to perform real-time analysis and store historical data;
[0016] The user platform is used to provide a user interface for vehicle owners and maintenance personnel.
[0017] As a further solution of the present invention, the specific steps of intelligent control of the monitoring management module are as follows:
[0018] Step 1: The monitoring management module measures the voltage, current and temperature of the battery in real time through built-in sensors, and calculates the battery's state of charge and health status based on the collected data. If the battery SOC is low, the owner is notified to charge to ensure that the battery has sufficient power. If the battery SOC is high, charging is postponed and the charging rate is reduced to avoid overcharging the battery.
[0019] Step 2: When the battery is charging, the monitoring and management module monitors the battery temperature, current, voltage, and charging rate in real time, and dynamically adjusts the charging current and voltage based on these data. When the battery is discharging, the monitoring and management module monitors the battery current and voltage in real time, and dynamically adjusts the discharge current and voltage according to the vehicle load demand to provide the required power;
[0020] Step 3: When the battery temperature or charging voltage exceeds the set threshold during the charging and discharging process, the monitoring and management module reduces the charging rate or stops charging, while recording the battery performance data in real time and generating a report for the owner or maintenance personnel to view.
[0021] As a further solution of the present invention, the specific calculation formula of the charge state in step 1 is as follows:
[0022]
[0023] In the formula, SOC(t) represents the battery state of charge at time t; SOC(0) represents the initial battery state of charge; Q represents the total capacity of the battery; I(τ) represents the value of the current at time τ;
[0024] The specific calculation formula of the health status described in step 1 is as follows:
[0025]
[0026] Where SOH stands for state of health.
[0027] As a further solution of the present invention, the load management module dynamically allocates power in the following specific steps:
[0028] Step 1: The load management module obtains the historical power demand data of different components of the vehicle, and obtains multiple sets of feature data after preprocessing the collected demand information. The acquired feature data is used to create new features through random replacement or sampling, and then different features are scaled and standardized to be on the same scale;
[0029] Step 2: Randomly divide each group of feature data into training set, test set and validation set, create a priority analysis model and determine the number of neurons in the model according to the required characteristics of each group of components, initialize the weight of each neuron to a random value or a preset initial value, and determine the model learning rate and step size according to the preset information;
[0030] Step 3: Divide the training set into small batches, pass each group of training set data to the priority analysis model in batches through forward propagation and calculate the output of each group of neurons, then use the loss function to compare the difference between the output of the priority analysis model and the actual target value, and then use the back propagation algorithm to calculate the gradient of the weight according to the gradient of the loss function, and update the weight of each neuron through the gradient descent algorithm;
[0031] Step 4: Use the validation set to evaluate the performance of the trained model. If the model performance does not meet the manager's expectations, adjust the model learning rate through cross-validation. When the model performance on the validation set no longer improves or begins to deteriorate, stop training early and evaluate the performance of the final priority analysis model through the test set. Repeat the model training and testing until all training sets have been used up, and then deploy the best performing neural network model to the control system.
[0032] Step 5: Regularly collect power demand data of different components of the vehicle through sensors or electronic system communication interfaces, pass the preprocessed data to the priority analysis model for forward propagation. The data starts from the input layer, passes through each hidden layer and calculates the weighted input of each layer of neurons. At the same time, the output of each layer of neurons is calculated through the activation function, and the output of the previous layer is passed to the next layer as input until it is passed to the output layer. After that, the output layer outputs the power demand ratio of different components of the vehicle according to the weight ratio of each neuron, and formulates an allocation strategy to allocate power from high to low according to the current demand.
[0033] As a further solution of the present invention, the specific steps of remote diagnosis by the fault diagnosis module are as follows:
[0034] Step ①: The fault diagnosis module periodically or continuously detects the operating status of the power system and collects information on various system parameters, while comparing each set of parameters with the predetermined normal operating range to identify any situation that exceeds or falls below the threshold;
[0035] Step 2: When an anomaly is detected, the system first uses the built-in fault database or knowledge base to compare the detected anomaly with known fault modes. When the fault is identified, the fault information is presented to the driver in the form of sound, vision or text.
[0036] Step 3: If no fault is identified, each set of abnormal information is serialized and mapped into a vector of fixed dimension. Then, a Bi-GRU network is constructed through a set of forward GRU and a set of reverse GRU, and multiple sets of attention heads are added to the Bi-GRU network.
[0037] Step ④: Splice multiple groups of attention heads to obtain the corresponding multi-head attention mechanism, and use multiple groups of attention heads to perform linear transformation on the same input from different angles to extract important features in the data required for air defense operations, fuse the outputs of multiple groups of attention heads, and input the fused feature vector into the subsequent layers;
[0038] Step ⑤: If the subsequent layer analyzes that there is a fault in the power system, the Bi-GRU network classifies the fault information through the classifier and generates a real-time alarm to notify the vehicle driver or maintenance personnel.
[0039] As a further solution of the present invention, the specific calculation formula of the multi-head attention mechanism in step ④ is as follows:
[0040]
[0041] MultiHead(Q,K,V)=Concat(head1,…,head h )W O (4);
[0042] where head = Attention (QW i Q ,KW i K ,VW i V ) (5);
[0043] Among them, W O , W i Q , W i K and W i Vrepresents the learnable parameter matrix in data projection, and h represents the number of attention heads.
[0044] As a further solution of the present invention, the specific steps of storing data in the record analysis module are as follows:
[0045] Step I: Segment each group of battery system data according to a preset time interval to form multiple groups of data blocks, then generate the identifier of each group of data blocks through a hash algorithm, collect each group of blockchain node information, and obtain the load status of each group of nodes;
[0046] Step II: Select the corresponding blockchain node to store each set of data blocks through the load balancing algorithm. After the data blocks are stored, the specified number of data blocks are copied to multiple sets of blockchain nodes according to the system requirements and available resources. When the data stored in the node changes, the data update is propagated from one node to other nodes through the data synchronization algorithm.
[0047] Step III: The record analysis module regularly loads the bytecode and metadata of each group of system data, uploads the data bytecode to the blockchain network, and assigns a unique contract address. After that, it constructs a new set of data blocks, broadcasts them to the blockchain network, and propagates them to multiple groups of blockchain nodes. Each group of blockchain nodes performs consensus verification, confirms the validity and legality of the data block, and adds it to the blockchain network for storage.
[0048] Compared with the prior art, the present invention has the following beneficial effects:
[0049] 1. The present invention obtains the power demand data of different parts of the historical vehicle, and obtains multiple groups of feature data after preprocessing the collected demand information. The acquired feature data is used to create new features through random replacement or sampling, and then different features are scaled and standardized to make them on the same scale. Each group of feature data is randomly divided into a training set, a test set, and a validation set, and a priority analysis model is created and optimized. The neural network model with the best performance is deployed to the control system, and the power demand data of different parts of the vehicle are regularly collected through sensors or electronic system communication interfaces, and the preprocessed data is passed to the priority analysis model for optimization. In forward propagation, data starts from the input layer, passes through each hidden layer and calculates the weighted input of each layer of neurons. At the same time, the output of each layer of neurons is calculated through the activation function, and the output of the previous layer is passed to the next layer as input until it is passed to the output layer. After that, the output layer outputs the proportion of power demand of different components of the vehicle according to the weight ratio of each neuron, and formulates an allocation strategy from high to low according to the current demand to allocate power. It can capture system behavior and demand changes from historical data, improve the accuracy of predicting future load demand, respond to changes in electronic system demand in a timely manner, improve the efficiency and accuracy of the load management module, and reduce the need for manual intervention.
[0050] 2. The present invention divides each group of battery system data according to a preset time interval to obtain multiple groups of data blocks, then generates an identifier of each group of data blocks through a hash algorithm, collects information of each group of blockchain nodes, and obtains the load status of each group of nodes. The corresponding blockchain node is selected through a load balancing algorithm to store each group of data blocks. After the data blocks are stored, a specified number of data blocks are copied to multiple groups of blockchain nodes according to system requirements and available resources. When the data stored in the node changes, the data update is propagated from one node to other nodes through a data synchronization algorithm. After a new group of data blocks is constructed, it is broadcast to the blockchain network and propagated to multiple groups of blockchain nodes. Each group of blockchain nodes performs consensus verification to confirm the validity and legitimacy of the data block and adds it to the blockchain network for storage, which can provide a high degree of data security, ensure the performance of the vehicle electronic system and the integrity of the fault data, establish trust among all participants and reduce information asymmetry, eliminate dependence on centralized data storage, and reduce the risk of single point failure. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.
[0052] Figure 1 This is a system block diagram of an MCU-based vehicle power control system proposed by the present invention. DETAILED DESCRIPTION
[0053] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.
[0054] Example 1
[0055] Reference Figure 1 , an MCU-based vehicle power control system, including a power input module, a converter, a monitoring management module, a load management module, an energy-saving decision module, a fault diagnosis module, a remote communication module, an energy integration module, a recording and analysis module and a user platform.
[0056] The power input module is used to receive power input from the vehicle battery and generator; the converter is used to convert the power from the battery and generator into a constant voltage and current; the monitoring and management module is used to monitor the battery status and intelligently control charging and discharging.
[0057] Specifically, the monitoring and management module measures the battery voltage, current and temperature in real time through built-in sensors, and calculates the battery charge state and health state based on the collected data sets. If the battery SOC is low, the owner is notified to charge to ensure sufficient battery power. If the battery SOC is high, charging is postponed and the charging rate is reduced to avoid overcharging the battery. When the battery is charging, the monitoring and management module monitors the battery temperature, current, voltage, and charging rate in real time, and dynamically adjusts the charging current and voltage based on these data. When the battery is discharging, the monitoring and management module monitors the battery current and voltage in real time, and dynamically adjusts the discharge current and voltage according to the vehicle's load requirements to provide the required electrical energy. When the battery temperature or charging voltage exceeds the set threshold during the charging and discharging process, the monitoring and management module reduces the charging rate or stops charging, and records the battery performance data in real time, and generates a report for the owner or maintenance personnel to view.
[0058] In this embodiment, the specific calculation formula of the charge state is as follows:
[0059]
[0060] In the formula, SOC(t) represents the battery state of charge at time t; SOC(0) represents the initial battery state of charge; Q represents the total capacity of the battery; I(τ) represents the value of the current at time τ;
[0061] The specific calculation formula for health status is as follows:
[0062]
[0063] Where SOH stands for state of health.
[0064] The load management module monitors the load requirements of each vehicle's electronic systems and dynamically allocates power; the energy-saving decision module is used to monitor the vehicle's energy efficiency and generate energy-saving decisions, while managing the start and stop of auxiliary systems.
[0065] Specifically, the load management module obtains historical power demand data of different components of the vehicle, and obtains multiple groups of feature data after preprocessing the collected demand information. New features are created by randomly replacing or sampling the acquired feature data, and then different features are scaled and standardized to be on the same scale. Each group of feature data is randomly divided into a training set, a test set, and a validation set. A priority analysis model is created and the number of neurons in the model is determined based on the demand characteristics of each group of components. At the same time, the weight of each neuron is initialized to a random value or a preset initial value, and the model learning rate and step size are determined based on the preset information. The training set is divided into small batches, and each group of training set data is passed to the priority analysis model in batches through forward propagation and the output of each group of neurons is calculated. The loss function is then used to compare the difference between the output of the priority analysis model and the actual target value. Then, the gradient of the weight is calculated based on the gradient of the loss function through the backpropagation algorithm, and the weight of each neuron is updated through the gradient descent algorithm. The validation set is used. To evaluate the performance of the trained model, if the model performance does not meet the expectations of the manager, the model learning rate is adjusted through the cross-validation method. When the performance of the model on the validation set no longer improves or begins to deteriorate, the training is stopped early, and the performance of the final priority analysis model is evaluated through the test set. The model is trained and tested repeatedly until all training sets are used up. The best performing neural network model is deployed to the control system. The power demand data of different components of the vehicle are regularly collected through sensors or electronic system communication interfaces, and the preprocessed data is passed to the priority analysis model for forward propagation. The data starts from the input layer, passes through each hidden layer and calculates the weighted input of each layer of neurons. At the same time, the output of each layer of neurons is calculated through the activation function, and the output of the previous layer is passed to the next layer as input until it is passed to the output layer. After that, the output layer outputs the power demand ratio of different components of the vehicle according to the weight ratio of each neuron, and formulates an allocation strategy to allocate power according to the current demand from high to low.
[0066] Example 2
[0067] Reference Figure 1 , an MCU-based vehicle power control system, including a power input module, a converter, a monitoring management module, a load management module, an energy-saving decision module, a fault diagnosis module, a remote communication module, an energy integration module, a recording and analysis module and a user platform.
[0068] The fault diagnosis module is used to detect power system faults to provide real-time alarms and remote diagnosis; the remote communication module is used to communicate with the vehicle's interconnected systems to receive remote commands and updates; and the energy integration module is used to integrate renewable energy to provide additional power.
[0069] Specifically, the fault diagnosis module periodically or continuously detects the operating status of the power system and collects information on various system parameters, and compares each group of parameters with a predetermined normal operating range to identify any situation that exceeds or falls below a threshold. When an abnormality is detected, the detected abnormality is first compared with a known fault mode using a built-in fault database or knowledge base. When the fault is identified, the fault information is presented to the driver in the form of sound, vision or text. If the fault is not identified, each group of abnormal information is serialized and mapped into a vector of fixed dimension. Then, a Bi-GRU network is constructed through a group of forward GRU and a group of reverse GRU, multiple groups of attention heads are added to the Bi-GRU network, multiple groups of attention heads are spliced, and corresponding multi-head attention mechanisms are obtained. The same input is linearly transformed from different angles through multiple groups of attention heads to extract important features from the data required for air defense operations, and the outputs of multiple groups of attention heads are fused, and the fused feature vector is input into the subsequent layer. If the subsequent layer analyzes that there is a fault in the power system, the Bi-GRU network classifies the fault information through a classifier and generates a real-time alarm to notify the vehicle driver or maintenance personnel.
[0070] Specifically, the calculation formula of the multi-head attention mechanism is as follows:
[0071]
[0072] MultiHead(Q,K,V)=Concat(head1,…,head h )W O (4);
[0073] where head = Attention (QW i Q ,KW i K ,VW i V ) (5);
[0074] Among them, W O , W i Q , W i K and W i V represents the learnable parameter matrix in data projection, and h represents the number of attention heads.
[0075] The recording and analysis module is used to record the power system performance data, and to perform real-time analysis and store historical data; the user platform is used to provide a user interface for vehicle owners and maintenance personnel.
[0076] Specifically, each group of battery system data is divided according to a preset time interval to form multiple groups of data blocks, and then the identifier of each group of data blocks is generated by a hash algorithm, the information of each group of blockchain nodes is collected, and the load of each group of nodes is obtained. The corresponding blockchain node is selected by the load balancing algorithm to store each group of data blocks. After the data blocks are stored, the specified number of data blocks are copied to multiple groups of blockchain nodes according to the system requirements and available resources. When the data stored in the node changes, the data update is propagated from one node to other nodes through the data synchronization algorithm. The record analysis module periodically loads the bytecode and metadata of each group of system data, uploads the data bytecode to the blockchain network, and assigns a unique contract address. After constructing a new group of data blocks, it is broadcast to the blockchain network and propagated to multiple groups of blockchain nodes. Each group of blockchain nodes performs consensus verification to confirm the validity and legality of the data block and adds it to the blockchain network for storage.
Claims
1. A vehicle power control system based on MCU, characterized in that: It includes power input module, converter, monitoring management module, load management module, energy-saving decision module, fault diagnosis module, remote communication module, energy integration module, record analysis module and user platform; The power input module is used to receive power input from the vehicle battery and the generator; The converter is used to convert power from the battery and the generator into a constant voltage and current; The monitoring and management module is used to monitor the battery status and intelligently control charging and discharging; The load management module monitors the load requirements of each vehicle electronic system and dynamically allocates power; The energy-saving decision module is used to monitor the energy efficiency of the vehicle and generate energy-saving decisions, while managing the start and stop of the auxiliary system; The fault diagnosis module is used to detect power system faults to provide real-time alarms and remote diagnosis; The remote communication module is used to communicate with the vehicle's interconnected systems and receive remote commands and updates; The energy integration module is used to integrate renewable energy to provide additional power; The recording and analysis module is used to record power system performance data, and to perform real-time analysis and store historical data; The user platform is used to provide a user interface for vehicle owners and maintenance personnel.
2. The MCU-based vehicle power control system according to claim 1, characterized in that: The specific steps of intelligent control of the monitoring management module are as follows: Step 1: The monitoring management module measures the voltage, current and temperature of the battery in real time through built-in sensors, and calculates the battery's state of charge and health status based on the collected data. If the battery SOC is low, the owner is notified to charge to ensure that the battery has sufficient power. If the battery SOC is high, charging is postponed and the charging rate is reduced to avoid overcharging the battery. Step 2: When the battery is charging, the monitoring and management module monitors the battery temperature, current, voltage, and charging rate in real time, and dynamically adjusts the charging current and voltage based on these data. When the battery is discharging, the monitoring and management module monitors the battery current and voltage in real time, and dynamically adjusts the discharge current and voltage according to the vehicle's load requirements to provide the required power; Step 3: When the battery temperature or charging voltage exceeds the set threshold during the charging and discharging process, the monitoring and management module reduces the charging rate or stops charging, records the battery performance data in real time, and generates a report for the owner or maintenance personnel to view.
3. The MCU-based vehicle power control system according to claim 2, characterized in that: The specific calculation formula of the charge state in step 1 is as follows: In the formula, SOC(t) represents the battery state of charge at time t; SOC(0) represents the initial battery state of charge; Q represents the total capacity of the battery; I(τ) represents the value of the current at time τ; The specific calculation formula of the health status described in step 1 is as follows: Where SOH stands for state of health.
4. The MCU-based vehicle power control system according to claim 2, characterized in that: The specific steps of dynamically allocating power by the load management module are as follows: Step 1: The load management module obtains the historical power demand data of different components of the vehicle, and obtains multiple sets of feature data after preprocessing the collected demand information. The acquired feature data is used to create new features through random replacement or sampling, and then different features are scaled and standardized to be on the same scale; Step 2: Randomly divide each group of feature data into training set, test set and validation set, create a priority analysis model and determine the number of neurons in the model according to the required characteristics of each group of components, initialize the weight of each neuron to a random value or a preset initial value, and determine the model learning rate and step size according to the preset information; Step 3: Divide the training set into small batches, pass each group of training set data to the priority analysis model in batches through forward propagation and calculate the output of each group of neurons, then use the loss function to compare the difference between the output of the priority analysis model and the actual target value, and then use the back propagation algorithm to calculate the gradient of the weight according to the gradient of the loss function, and update the weight of each neuron through the gradient descent algorithm; Step 4: Use the validation set to evaluate the performance of the trained model. If the model performance does not meet the manager's expectations, adjust the model learning rate through cross-validation. When the model performance on the validation set no longer improves or begins to deteriorate, stop training early and evaluate the performance of the final priority analysis model through the test set. Repeat the model training and testing until all training sets have been used up, and then deploy the best performing neural network model to the control system. Step 5: Regularly collect power demand data of different components of the vehicle through sensors or electronic system communication interfaces, pass the preprocessed data to the priority analysis model for forward propagation. The data starts from the input layer, passes through each hidden layer and calculates the weighted input of each layer of neurons. At the same time, the output of each layer of neurons is calculated through the activation function, and the output of the previous layer is passed to the next layer as input until it is passed to the output layer. After that, the output layer outputs the power demand ratio of different components of the vehicle according to the weight ratio of each neuron, and formulates an allocation strategy to allocate power from high to low according to the current demand.
5. The MCU-based vehicle power control system according to claim 1, characterized in that: The specific steps of remote diagnosis of the fault diagnosis module are as follows: Step ①: The fault diagnosis module periodically or continuously detects the operating status of the power system and collects information on various system parameters, while comparing each set of parameters with the predetermined normal operating range to identify any situation that exceeds or falls below the threshold; Step 2: When an anomaly is detected, the system first uses the built-in fault database or knowledge base to compare the detected anomaly with known fault modes. When the fault is identified, the fault information is presented to the driver in the form of sound, vision or text. Step 3: If no fault is identified, each set of abnormal information is serialized and mapped into a vector of fixed dimension. Then, a Bi-GRU network is constructed through a set of forward GRU and a set of reverse GRU, and multiple sets of attention heads are added to the Bi-GRU network. Step ④: Splice multiple groups of attention heads to obtain the corresponding multi-head attention mechanism, and use multiple groups of attention heads to perform linear transformation on the same input from different angles to extract important features in the data required for air defense operations, fuse the outputs of multiple groups of attention heads, and input the fused feature vector into the subsequent layers; Step ⑤: If the subsequent layer analyzes that there is a fault in the power system, the Bi-GRU network classifies the fault information through the classifier and generates a real-time alarm to notify the vehicle driver or maintenance personnel.
6. The MCU-based vehicle power control system according to claim 5, characterized in that: The specific calculation formula of the multi-head attention mechanism in step ④ is as follows: MultiHead(Q,K,V)=Concat(head1,…,head h )W O (4); wherehead =Attention (QW i Q ,KW i K ,VW i V ) (5); Among them, W O , W i Q , W i K and W i V represents the learnable parameter matrix in data projection, and h represents the number of attention heads.
7. The MCU-based vehicle power control system according to claim 1, characterized in that: The specific steps of data storage of the record analysis module are as follows: Step I: Segment each group of battery system data according to a preset time interval to form multiple groups of data blocks, then generate the identifier of each group of data blocks through a hash algorithm, collect each group of blockchain node information, and obtain the load status of each group of nodes; Step II: Select the corresponding blockchain node to store each set of data blocks through the load balancing algorithm. After the data blocks are stored, the specified number of data blocks are copied to multiple sets of blockchain nodes according to the system requirements and available resources. When the data stored in the node changes, the data update is propagated from one node to other nodes through the data synchronization algorithm. Step III: The record analysis module regularly loads the bytecode and metadata of each group of system data, uploads the data bytecode to the blockchain network, and assigns a unique contract address. After that, it constructs a new set of data blocks, broadcasts them to the blockchain network, and propagates them to multiple groups of blockchain nodes. Each group of blockchain nodes performs consensus verification, confirms the validity and legality of the data block, and adds it to the blockchain network for storage.
Citation Information
Patent Citations
Method for achieving dynamic balance of automobile electric quantity and automobile electronic controller
CN103855751A
Lithium battery control system and control method thereof
CN116961186A
Intelligent decision-making method for air defense combat based on potential game and deep reinforcement learning
CN117057233A
Electrical equipment preventive maintenance optimization method and device based on reliability
CN117391674A
Distribution system of DAPPS of block chain
CN117609372A
Cited By
Electric power electronic control distribution system
CN120382787A
