Control method for heating assembly of electronic thermostat of hydrogen fuel engine based on Internet of Things

Abstract

The invention discloses a hydrogen fuel engine electronic thermostat heating assembly control method based on the Internet of Things, particularly relates to the technical field of hydrogen fuel engine control, and is used for solving the problems of cooling response lag and action conflict caused by disconnection of control logic and a real-time network state of an existing temperature control system in a dynamic network environment. Electronic thermostat temperature data and network state data are collected in real time to generate a dynamic synchronization coefficient, network topology toughness evaluation and causal influence analysis are combined to predict instruction conflict areas and divide priorities, and gradient correction is performed on heating power and valve opening based on a target deviation compensation value. And a correction instruction is issued through the distributed control unit and a communication reliability weight is synchronously updated, so that dynamic coordination of a network state and a temperature control parameter is realized, and efficient and stable operation of the hydrogen fuel engine under a complex working condition is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrogen fuel engine control, and more specifically, to a control method for an electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things. Background Art

[0002] Hydrogen fuel engines have the advantages of zero carbon emissions and high energy conversion efficiency. To ensure their efficient and stable operation, it is necessary to coordinate the control of the cooling system temperature through an electronic thermostat and a heating assembly, such as rapid heating during low-temperature cold start and precise heat dissipation under high-temperature conditions. Currently, the temperature control system of hydrogen fuel engines mostly relies on local sensors and controllers to achieve closed-loop regulation, and the introduction of Internet of Things technology makes cloud collaborative control possible, such as optimizing the heating power distribution through remote data interaction. However, in the prior art, the temperature control architecture based on the Internet of Things is still limited to the design of a single communication link or static node relationships, and does not fully consider complex problems such as dynamic access of network nodes and asynchronous data timing in actual scenarios.

[0003] In the existing Internet-of-Things-based temperature control method for hydrogen fuel engines, in a dynamic heterogeneous network environment, due to the disconnection between the collaborative control logic of the electronic thermostat and the heating assembly and the real-time network state, the response of the cooling system lags or there are action conflicts, making it difficult for the hydrogen fuel engine to maintain thermal balance under complex working conditions, directly affecting the combustion efficiency and component life. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a control method for an electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A control method for an electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things, comprising: S1. Real-time collect the temperature data of the electronic thermostat of the hydrogen fuel engine and the network state data of the Internet of Things nodes, where the network state data includes dynamic access node identifiers and packet transmission delays; S2. Generate a dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamic access node identifier and the packet transmission delay; S3. Based on the dynamic synchronization coefficient and the causal influence in the network topology corresponding to the dynamic access node identifier, predict the spatio-temporal conflict region of the multi-node control instruction and divide the conflict priority; S4. When the dynamic synchronization coefficient is lower than a preset synchronization threshold, analyze the network topology toughness of the dynamic access node identifier, and generate a target deviation compensation value in combination with the electronic thermostat temperature data and the conflict priority; S5. Gradient-correct the total heating success rate command according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat; S6. Send the corrected total heating success rate command and the valve opening reference parameter through the distributed control unit of the Internet of Things node, and update the communication reliability weight.

[0006] In a preferred embodiment, the temperature data of the electronic thermostat of the hydrogen fuel engine and the network status data of the Internet of Things node are collected in real time, including: Obtain the temperature data of the electronic thermostat through the vehicle-mounted communication terminal; Based on the vehicle kinematic parameters corresponding to the dynamically accessed node identifier, parse the network status data of the Internet of Things node in real time; Calculate the data packet transmission delay by the time stamp difference method; Store the temperature data of the electronic thermostat, the dynamically accessed node identifier, and the data packet transmission delay in the distributed cache queue.

[0007] In a preferred embodiment, generate the dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamically accessed node identifier and the data packet transmission delay, including: Normalize the communication reliability weight of the dynamically accessed node identifier; Generate a dynamic delay correction factor according to the stability evaluation result of the data packet transmission delay. The stability evaluation is achieved by calculating the ratio of the variance to the mean of the data packet transmission delays of a continuous preset number of data packets; Input the normalized communication reliability weight and the dynamic delay correction factor into the weighted fusion function to generate the dynamic synchronization coefficient. The weighted fusion function dynamically adjusts the fusion ratio of the normalized communication reliability weight and the dynamic delay correction factor based on the current network protocol type; Update the communication reliability weight according to the real-time change rate of the dynamic synchronization coefficient.

[0008] In a preferred embodiment, based on the dynamic synchronization coefficient and the causal influence in the network topology corresponding to the dynamically accessed node identifier, predict the spatio-temporal conflict area of the multi-node control command and divide the conflict priority, including: Extract the real-time stability parameter of the control link corresponding to the dynamic synchronization coefficient; Construct a causal influence map according to the communication path dependence of the dynamically accessed node identifier; Analyze the spatio-temporal attribute distribution of the node commands in the causal influence map through the spatio-temporal overlap detection algorithm to identify the spatio-temporal conflict area; Generate a conflict priority ranking table based on the real-time stability parameter and the historical conflict probability in the spatio-temporal conflict area. The ranking rule of the conflict priority ranking table is to assign the highest conflict priority to the area where the real-time stability parameter is lower than the preset stability threshold and the historical conflict probability is higher than the preset probability threshold; Dynamically match the conflict priority ranking table with the frequency response characteristics of the dynamic synchronization coefficient, and adjust the priority division threshold according to the periodic and burst characteristics of the control instruction.

[0009] In a preferred embodiment, the real-time stability parameter is generated through the collaborative analysis of the historical volatility of the dynamic synchronization coefficient and the current network load state. The historical volatility is the standard deviation of the dynamic synchronization coefficient in the past preset time period, and the current network load state is the ratio of the number of data packets received by the vehicle node per unit time to the total bandwidth.

[0010] In a preferred embodiment, the communication path dependence is generated by the non-linear combination of the covariance matrix of the historical communication success rate between nodes and the path redundancy coefficient. The path redundancy coefficient is the ratio of the number of available communication paths between nodes to the total number of paths.

[0011] In a preferred embodiment, when the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier, and generate a target deviation compensation value in combination with the electronic thermostat temperature data and the conflict priority, including: When the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier. The network topology resilience is generated by the non-linear association of the path redundancy coefficient and the node connectivity; Generate an initial compensation coefficient according to the mapping relationship between the real-time deviation rate of the electronic thermostat temperature data and the conflict priority; Adjust the initial compensation coefficient through the protocol type adaptive function. The protocol type adaptive function scales the initial compensation coefficient according to the network protocol type of the dynamically accessed node identifier. The scaling factor under the Transmission Control Protocol is the reciprocal of the path redundancy coefficient; Input the adjusted initial compensation coefficient and the network topology resilience into the deviation fusion model to generate the target deviation compensation value.

[0012] In a preferred embodiment, perform gradient correction on the total heating success rate instruction according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat, including: Generate a gradient correction step size based on the current change rate and the historical change trend of the target deviation compensation value; Adjust the correction amplitude according to the gradient correction step size and the deviation direction of the current total heating success rate instruction. The deviation direction is determined by the positive and negative signs of the target deviation compensation value; Dynamically match the corrected total heating power command with the real-time temperature feedback value of the electronic thermostat. If the corrected command exceeds the preset heating power range, it is forced to be truncated to the boundary value through a clipping function; Reverse-derive the reference parameter adjustment amount of the electronic thermostat valve opening according to the corrected total heating power command; Eliminate the instantaneous jitter of the adjustment amount through the hysteresis compensation function of the valve opening controller, and write the updated valve opening reference parameter into the control command queue and send it to the actuator.

[0013] In a preferred embodiment, the corrected total heating power command and the valve opening reference parameter are sent by the distributed control unit of the Internet of Things node, and the communication reliability weight is updated, including: Generate a multicast transmission strategy based on the priority levels of the commands in the conflict priority sorting table and the network topology resilience parameters; Select the instruction encoding format according to the network protocol type and the real-time load status. The segmented acknowledgment retransmission mechanism is adopted under the Transmission Control Protocol, and the redundant data packet broadcast mechanism is adopted under the User Datagram Protocol; Schedule the instruction sending order through the queue management module of the distributed control unit. The scheduling rule is that high-priority commands in the conflict priority sorting table are preferentially allocated high-bandwidth resources, and low-priority commands are delayed for sending; Update the communication reliability weight after receiving the instruction confirmation signal from the actuator. The update logic is to dynamically adjust the weight gain coefficient based on the dynamic feedback of the confirmation signal response time and the packet loss rate; Synchronize the updated communication reliability weight to the node registry of the cloud server, and broadcast it to associated nodes through the heartbeat packet mechanism to maintain topological consistency.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. Through the dynamic coordination of the electronic thermostat temperature data and the network status data, the temperature control system of the hydrogen fuel engine realizes the adaptive adjustment to the complex network environment. Based on the communication reliability weight and the real-time transmission delay of the dynamically accessed nodes, a dynamic synchronization coefficient is generated, enabling the control link to accurately perceive network fluctuations and dynamically adjust the response strategy; by analyzing the causal influence to predict the spatio-temporal conflict area of multi-node instructions and combining the generation logic of the network topology resilience evaluation compensation value, it is ensured that the temperature deviation can still be quickly corrected when the link is unstable. Solve the problem of response lag caused by the disconnection between the network state and the temperature control logic in the traditional method, so that the cooling system can maintain thermal balance in scenarios such as low-temperature cold start and high-load heat dissipation, significantly improving the combustion efficiency and the service life of components.

[0015] 2. A closed-loop feedback is formed through the gradient correction mechanism and distributed instruction issuance, further optimizing the real-time performance and reliability of control instructions. Based on the target deviation compensation value, the coordinated adjustment of the heating power and valve opening avoids overshoot or oscillation caused by single-parameter correction; the dynamically updated communication reliability weight and network topology resilience parameter form a dynamic mapping, enabling the distributed control unit to adaptively allocate resources according to the node status. While ensuring the temperature control accuracy, the risk of instruction conflicts in heterogeneous network environments is effectively reduced, providing a reliable guarantee for the stable operation of the hydrogen fuel engine under complex working conditions. Brief Description of the Drawings

[0016] Figure 1 It is a flowchart of the control method for the electronic thermostat heating assembly of the hydrogen fuel engine based on the Internet of Things according to the present invention; Figure 2 It is a flowchart of predicting the spatio-temporal conflict area and dividing the conflict priority according to the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment: Figure 1 The control method for the electronic thermostat heating assembly of the hydrogen fuel engine based on the Internet of Things according to the present invention is given, including: S1. Real-time collect the temperature data of the electronic thermostat of the hydrogen fuel engine and the network status data of the Internet of Things nodes, where the network status data includes the dynamically accessed node identifier and the data packet transmission delay; S2. Generate the dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamically accessed node identifier and the data packet transmission delay; S3. Based on the dynamic synchronization coefficient and the causal influence in the network topology corresponding to the dynamically accessed node identifier, predict the spatio-temporal conflict area of the multi-node control instruction and divide the conflict priority; S4. When the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier, and generate the target deviation compensation value in combination with the electronic thermostat temperature data and the conflict priority; S5. Gradient-correct the heating total power instruction according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat; S6. Send the corrected total heating power instruction and the valve opening reference parameter through the distributed control unit of the Internet of Things node, and update the communication reliability weight.

[0019] S1. Real-time collect the electronic thermostat temperature data of the hydrogen fuel engine and the network status data of the Internet of Things node. The network status data includes the dynamic access node identifier and the data packet transmission delay. The specific implementation is as follows: Obtain the electronic thermostat temperature data of the hydrogen fuel engine through the vehicle-mounted communication terminal. The electronic thermostat temperature data includes the coolant inlet temperature and the coolant outlet temperature. The coolant inlet temperature is collected by a PT100 platinum resistance temperature sensor installed at the coolant inlet of the engine block, with a measurement range of -40°C to 150°C. The coolant outlet temperature is collected by a K-type thermocouple temperature sensor installed on the radiator return pipe, with a measurement range of 0°C to 200°C. Both temperature sensors are calibrated according to the IEC 60751 standard. After converting the real-time temperature data into digital signals, the temperature sensors transmit the data to the vehicle-mounted communication terminal through the Controller Area Network bus. The communication rate of the Controller Area Network bus is 250 kbps, and the data frame format conforms to the ISO 11898-2 standard. The vehicle-mounted communication terminal encapsulates the temperature data into a data packet conforming to the MQTT Internet of Things communication protocol, and the payload part of the data packet is encoded in JSON format.

[0020] The network status data of the Internet of Things node is collected in real time through the wireless communication links between the vehicle-mounted communication terminal and the cloud server, the roadside unit, and other vehicle nodes. The network status data includes the dynamic access node identifier and the data packet transmission delay. The dynamic access node identifier is composed of the binding relationship between the vehicle unique code and the network address. The vehicle unique code is a 32-byte hexadecimal string generated by applying the SHA-256 hash algorithm to the last 8 characters of the vehicle identification number. The network address is a temporary Internet protocol address assigned by the vehicle-mounted communication terminal in the vehicle-to-everything communication protocol according to the IEEE 802.11p standard, and the address allocation period is 10 minutes. The binding relationship between the vehicle unique code and the network address is dynamically updated through the OAuth 2.0 identity authentication service of the cloud server to ensure the uniqueness of the node identity and the traceability of the communication link.

[0021] Based on the vehicle kinematic parameters corresponding to the dynamic access node identifier, the network status data of the Internet of Things node is parsed in real time. The vehicle kinematic parameters include the vehicle's real-time position coordinates, driving speed, and acceleration. The vehicle's real-time position coordinates are obtained through an in-vehicle global positioning system with a positioning accuracy of ±1.5 meters and a data update frequency of 10Hz. The driving speed and acceleration are collected through an in-vehicle six-axis inertial measurement unit. The range of the inertial measurement unit is ±200° / s (angular velocity) and ±16g (acceleration), and the sampling frequency is 100Hz. The binding relationship between the vehicle kinematic parameters and the dynamic access node identifier is implemented through the header fields of the WSMP data frame of the vehicle networking communication protocol. The WSMP data frame header fields include the dynamic access node identifier, UTC timestamp, and vehicle kinematic parameter encoding. The timestamp format is YYYY-MM-DD HH:MM:SS.sss, and the vehicle kinematic parameter encoding uses the ASN.1 PER compressed format. The process of parsing the network status data of the Internet of Things node includes extracting the dynamic access node identifier from the WSMP data frame header and querying the node registry of the cloud server according to the dynamic access node identifier. The storage fields of the node registry include the communication reliability weight and historical connection record of the vehicle node. The communication reliability weight is dynamically calculated based on the historical packet transmission success rate, connection duration, and disconnection recovery speed of the vehicle node in the past 30 minutes. The historical packet transmission success rate is the ratio of the number of successfully received packets to the total number of sent packets. The connection duration is the duration of a single communication link (unit: second), and the disconnection recovery speed is the average time for the node to re-establish a connection after disconnection (unit: millisecond). The historical connection record includes the communication frequency and data traffic of the vehicle node with other nodes in the past 1 hour. The communication frequency statistical period is per minute, and the data traffic statistical unit is megabyte.

[0022] Calculate the packet transmission delay through the timestamp difference method. The timestamp difference method includes recording the clock difference between the packet sending time and the receiving time, and dynamically correcting the clock offset error according to the network protocol type. The packet sending time is written into the timestamp field of the WSMP data frame header by the sending vehicle communication terminal when encapsulating the packet, and the timestamp accuracy is at the millisecond level. The packet receiving time is recorded by the receiving vehicle communication terminal's local system clock when parsing the packet. The local system clock is synchronized with the UTC time through the NTP protocol. The clock difference is calculated by taking the absolute value of the receiving time minus the sending time. The correction of the clock offset error adopts different strategies according to the network protocol type. For example, in Transmission Control Protocol communication, the clock deviation between the sending and receiving ends is calibrated through the sequence number and acknowledgment mechanism. The sequence number matching rule is that the receiving end performs a linear regression fit on the sequence numbers of 3 consecutive packets to estimate the one-way propagation delay. In User Datagram Protocol communication, the clock offset is dynamically estimated through the round-trip time of the heartbeat packet at the application layer. The heartbeat packet sending interval is 5 seconds, and the round-trip time calculation formula is RTT = receiving time - sending time. The corrected packet transmission delay parameter is filtered by the sliding window average algorithm to filter out instantaneous jitters. The sliding window size is 10 packets, and the data within the window is weighted and averaged in chronological order. The weight coefficient is an exponential decay function with a value of 0.9 to obtain a stable delay estimate in milliseconds.

[0023] Store the electronic thermostat temperature data, dynamic access node identifier, and packet transmission delay parameter in the distributed cache queue. The distributed cache queue allocates storage resources based on the priority of the network status data. The priority is dynamically adjusted according to the real-time requirements of the data type. For example, the coolant inlet temperature and outlet temperature data, which directly affect the engine thermal management decision, are assigned the highest priority and stored in the low-latency memory cache area. The memory cache area has a capacity of 128MB and adopts the LRU replacement strategy. The dynamic access node identifier and packet transmission delay parameter, which are used for network status evaluation, are assigned medium priority and stored in the high-speed solid-state drive cache area. The solid-state drive cache area has a capacity of 1GB and adopts the FIFO replacement strategy. The data writing and reading operations of the distributed cache queue are implemented through multi-threaded concurrent control. The thread pool size is twice the number of CPU cores. The mutex mechanism is implemented using pthread_mutex of the POSIX thread library to ensure the timeliness of high-priority data processing. The storage resource allocation strategy of the cache queue is dynamically optimized through the Kubernetes resource scheduling service of the cloud server, and the occupancy ratio of memory and storage media is adjusted according to the real-time network load and computing power of the vehicle node. For example, when the network is congested (packet loss rate exceeds 5%), the memory cache area is temporarily expanded to 256MB to ensure the real-time processing ability of temperature data. The expansion trigger condition is that the packet loss rate threshold is exceeded in 3 consecutive sampling periods.

[0024] S2. Generate the dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamic access node identifier and the packet transmission delay. The specific implementation is as follows: Normalize the communication reliability weight of the dynamic access node identifier. The normalization process calculates the weight scaling coefficient based on the sliding average and standard deviation of the historical connection success rate within a preset time window. The duration of the preset time window is set through the configuration parameters of the cloud server, with a default value of 10 minutes. The vehicle node counts all communication connection attempts and successful attempts within 10 minutes. The calculation method of the historical connection success rate is the number of successful attempts divided by the total number of attempts and converted to a percentage. For example, if 100 connections are initiated within 10 minutes and 90 are successful, the historical connection success rate is 90%. The sliding average is calculated using the time-weighted exponential moving average method, and the time decay coefficient is dynamically adjusted according to the network load, with a default value of 0.9. The specific implementation is that the current sliding average is equal to the previous sliding average multiplied by the time decay coefficient plus the current historical connection success rate multiplied by (1 minus the time decay coefficient). The standard deviation is calculated by taking the square root of the mean of the sum of the squared deviations between the sliding average and the historical connection success rate. The sum of squared deviations is the cumulative sum of the squares of the differences between each historical connection success rate and the sliding average within 10 minutes. The mean is the sum of squared deviations divided by the number of data points, and the standard deviation is the square root of the mean. The calculation method of the weight scaling coefficient is the communication reliability weight divided by the product of the standard deviation and the sliding average. If the calculation result exceeds 1, the weight scaling coefficient is forced to be set to 1. If the calculation result is less than 0, the weight scaling coefficient is forced to be set to 0. For example, if the communication reliability weight is 80, the standard deviation is 5, and the sliding average is 80, the weight scaling coefficient is 80 divided by (5×80) which equals 0.25, and after clipping, it takes 0.25.

[0025] Generate the dynamic delay correction factor according to the stability evaluation result of the packet transmission delay. The stability evaluation is achieved by calculating the ratio of the variance to the mean of the transmission delays of a continuous preset number of packets. The continuous preset number is dynamically determined by the computing resources of the vehicle node, with a default value of 30 packets. The packet transmission delay sequence is stored in a circular buffer in chronological order. The mean is the arithmetic average of the 30 delay values, and the variance is the mean of the sum of the squares of the differences between each delay value and the mean. The calculation method of the ratio of variance to mean is variance divided by mean. If the mean is 0 or less than 1 millisecond, the mean is forced to be set to 1 millisecond to avoid division by zero errors. The generation method of the dynamic delay correction factor is the reciprocal of the ratio of variance to mean multiplied by a preset proportionality coefficient. The preset proportionality coefficient is dynamically configured according to the network protocol type. The default proportionality coefficient for the Transmission Control Protocol is 0.5, and the default proportionality coefficient for the User Datagram Protocol is 0.3. The calculation result of the dynamic delay correction factor is constrained between 0 and 5 through a clipping function. If the calculation result exceeds 5, it is set to 5. If the calculation result is less than 0, it is set to 0.

[0026] The normalized communication reliability weight and the dynamic delay correction factor are input into the weighted fusion function to generate a dynamic synchronization coefficient. The weighted fusion function dynamically adjusts the fusion ratio of the normalized communication reliability weight and the dynamic delay correction factor based on the current network protocol type. The calculation method of the weighted fusion function is the normalized communication reliability weight multiplied by the weight ratio plus the dynamic delay correction factor multiplied by the factor ratio, and the calculation result is converted into an integer range from 0 to 100 through linear mapping. The mapping method is to round the original calculation result multiplied by 100 to the nearest integer. If the calculation result exceeds 100, it is forced to be set to 100. If the calculation result is less than 0, it is forced to be set to 0.

[0027] In the Transmission Control Protocol, the proportion of the normalized communication reliability weight is configured as 60%, and the proportion of the dynamic delay correction factor is configured as 40%. In the User Datagram Protocol, the proportion of the normalized communication reliability weight is configured as 40%, and the proportion of the dynamic delay correction factor is configured as 60%. The setting basis is as follows: The Transmission Control Protocol is a connection-oriented reliable transmission protocol, and its communication reliability weight reflects the stability of the long-term connection between nodes. When there is network congestion or delay fluctuation, the reliability needs to be guaranteed first, so a higher weight is assigned; the User Datagram Protocol is a connectionless real-time transmission protocol, and the transmission delay of its data packets directly affects the real-time performance of control instructions, so a higher weight of the dynamic delay correction factor is assigned to optimize the real-time response. The proportional coefficient is verified by experimental data. In a typical vehicle networking scenario, when the Transmission Control Protocol adopts a 60% reliability weight, the disconnection rate is reduced by 30%. When the User Datagram Protocol adopts a 60% delay correction factor, the instruction response delay is reduced by 25%. And it is allowed to dynamically adjust the proportional coefficient according to the vehicle node type (such as the host vehicle / following vehicle). For example, an emergency vehicle can temporarily increase the proportion of the delay correction factor to 70%.

[0028] Update the communication reliability weight according to the real-time change rate of the dynamic synchronization coefficient. The real-time change rate is calculated by the ratio of the absolute value of the difference between the current dynamic synchronization coefficient and the previous dynamic synchronization coefficient to the time interval. The time interval is the difference between the current moment and the previous calculation moment, in seconds. The calculation period of the real-time change rate is dynamically adjusted by the control frequency of the vehicle node, with a default value of once every 5 seconds. The update rule of the communication reliability weight is as follows: if the real-time change rate exceeds 5 units per second, the current communication reliability weight is multiplied by a decay coefficient of 0.9; if the real-time change rate is less than 1 unit per second, the current communication reliability weight is multiplied by a gain coefficient of 1.1; if the real-time change rate is between 1 unit and 5 units per second, the communication reliability weight remains unchanged. The updated communication reliability weight is rounded to the nearest integer and restricted to the range of 0 to 100. If the calculation result exceeds 100, it is forced to be set to 100. If the calculation result is less than 0, it is forced to be set to 0. The update result is synchronized to the node registry of the cloud server. The fields stored in the node registry include the vehicle's unique code, network address, communication reliability weight, and historical connection records, which are used for subsequent control link calls.

[0029] Figure 2 The flowchart of predicting the spatio-temporal conflict area and dividing the conflict priority of the present invention is given. S3. Based on the causal influence in the network topology corresponding to the dynamic synchronization coefficient and the dynamic access node identifier, predict the spatio-temporal conflict area of the multi-node control instruction and divide the conflict priority. The specific implementation is as follows: Extract the real-time stability parameter of the control link corresponding to the dynamic synchronization coefficient. The real-time stability parameter is generated through the collaborative analysis of the historical volatility of the dynamic synchronization coefficient and the current network load status. The historical volatility of the dynamic synchronization coefficient is the standard deviation of the dynamic synchronization coefficient in the past 10 minutes. The calculation method of the standard deviation is the square root of the mean of the sum of the squares of the differences between each dynamic synchronization coefficient and the mean of the dynamic synchronization coefficients in 10 minutes. The current network load status is the ratio of the number of data packets received by the vehicle node per unit time to the total bandwidth. The unit time is set to 1 second, and the total bandwidth is determined according to the communication module specifications of the vehicle node. For example, the total bandwidth of an in-vehicle LTE module is 100 Mbps. If a vehicle node receives 1000 data packets in 1 second and the average size of each data packet is 1 KB, then the current network load status is (1000×8 Kb)÷100 Mbps = 0.08. The generation method of the real-time stability parameter is the quotient of the historical volatility divided by the current network load status. If the quotient exceeds 10, it is forced to be set to 10. If the quotient is less than 0, it is forced to be set to 0.

[0030] A causal influence graph is constructed based on the communication path dependencies of dynamically accessed node identifiers. Communication path dependencies are generated by a nonlinear combination of the covariance matrix of historical communication success rates between nodes and the path redundancy coefficient. The historical communication success rate is the ratio of the number of successful communication connections between two nodes in the past hour to the total number of attempts. The covariance matrix is calculated as a symmetric matrix consisting of the covariance values of the historical communication success rates of all node pairs. The covariance value reflects the linkage between the communication success rates between nodes. For example, if the communication success rates of nodes A and B are 80% and 70%, respectively, a positive covariance value indicates a positive correlation in communication stability between the two nodes. The path redundancy coefficient is the ratio of the number of available communication paths between nodes to the total number of paths. The total number of paths is determined by the IoV topology. For example, if there are three communication paths from node A to node B, two of which are direct paths and one is a relay path, the number of available communication paths is 3. The total number of paths is set to 5 based on the maximum number of paths allowed by the network protocol. The path redundancy coefficient is 3 divided by 5, which equals 0.6. The nonlinear combination is generated by multiplying each element of the covariance matrix by the path redundancy coefficient of the corresponding node pair, and then normalizing the result to a range of -1 to 1.

[0031] A spatiotemporal overlap detection algorithm analyzes the spatiotemporal attribute distribution of node instructions in the causal influence graph to identify spatiotemporal conflict areas. These spatiotemporal attributes include the start and end times of the instruction execution window and the Euclidean distance offset of the vehicle's position. The start and end times of the instruction execution window are determined by the control instruction's generation time and the preset execution cycle. For example, if a heating total success rate instruction is generated at 10:00:00 and has a 5-second execution cycle, the window is from 10:00:00 to 10:00:05. The Euclidean distance offset of the vehicle's position is the straight-line distance between the vehicle's position and the target position at the time of instruction execution. It is calculated using Global Positioning System coordinates. For example, if the vehicle's current latitude is 39.9 degrees and its longitude is 116.3 degrees, and the target position is 40.0 degrees and its longitude is 116.4 degrees, the Euclidean distance offset is approximately 11.1 kilometers. The logic of the spatiotemporal overlap detection algorithm is that if the time windows of two instructions intersect and the Euclidean distance offset of the vehicle position is less than a preset threshold (for example, 100 meters), it is determined to be a spatiotemporal conflict area.

[0032] Generate a conflict priority ranking table based on real-time stability parameters and historical conflict probabilities in the spatio-temporal conflict area. The historical conflict probability is the percentage of the number of control instruction conflicts that occurred in a certain area within the past 24 hours in the total number of instructions. For example, if a certain area executed 1000 instructions within the past 24 hours and 50 of them had conflicts, then the historical conflict probability is 5%. The sorting rule of the priority ranking table is that areas with real-time stability parameters lower than the preset stability threshold (e.g., 5) and historical conflict probabilities higher than the preset probability threshold (e.g., 3%) are assigned the highest conflict priority. The preset stability threshold and probability threshold are dynamically adjusted through experimental data. For example, when the network load is higher than 50%, the preset stability threshold is reduced to 3 to increase sensitivity.

[0033] Dynamically match the conflict priority ranking table with the frequency response characteristics of the dynamic synchronization coefficient, and adjust the priority division threshold according to the periodic and bursty characteristics of the control instructions. The frequency response characteristics of the dynamic synchronization coefficient are the fluctuation amplitude of the dynamic synchronization coefficient at different control instruction frequencies. For example, the fluctuation amplitude of the dynamic synchronization coefficient corresponding to high-frequency instructions (10 times per second) is 5, and the fluctuation amplitude corresponding to low-frequency instructions (1 time per second) is 1. The adjustment method of the priority division threshold is to reduce the threshold of high-frequency instructions by 20% to reduce resource competition conflicts, and increase the threshold of low-frequency instructions by 10% to allow a greater tolerance for fluctuations. The adjusted threshold is sent to the vehicle nodes through the policy engine of the cloud server to ensure that all nodes adopt a unified conflict priority division standard.

[0034] S4. When the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier, and generate a target deviation compensation value in combination with the electronic thermostat temperature data and the conflict priority. The specific implementation is as follows: When the dynamic synchronization coefficient is lower than the preset synchronization threshold, parse the network topology resilience of the dynamically accessed node identifier. The network topology resilience is generated through the non-linear association of the path redundancy coefficient and the node connectivity. The path redundancy coefficient is the ratio of the number of available communication paths to the total number of paths. The total number of paths is determined by the maximum number of paths allowed by the vehicle network communication protocol. For example, in a vehicular ad hoc network, the maximum number of paths from node A to node B is 5. If the current available paths are 3, then the path redundancy coefficient is 3 divided by 5, which is equal to 0.6. The node connectivity is the number of nodes that the dynamically accessed node directly communicates with in the current topology. For example, if node C directly connects to node D, node E, and node F in the vehicle network, then the node connectivity is 3. The generation method of the network topology resilience is the path redundancy coefficient multiplied by the natural logarithm of the node connectivity. For example, if the path redundancy coefficient is 0.6 and the node connectivity is 3, then the network topology resilience is 0.6×ln(3)≈0.6×1.0986≈0.659. The calculation result of the network topology resilience is constrained between 0 and 1 through a clipping function. If the calculation result exceeds 1, it is forced to be set to 1. If the calculation result is less than 0, it is forced to be set to 0.

[0035] Generate an initial compensation coefficient according to the mapping relationship between the real-time deviation rate of the electronic thermostat temperature data and the conflict priority. The real-time deviation rate is the percentage of the absolute value of the difference between the current temperature and the target temperature in the target temperature. For example, if the target temperature is 80 °C and the current temperature is 75 °C, the real-time deviation rate is |75 - 80| ÷ 80 × 100% = 6.25%. The conflict priority is determined by the conflict priority sorting table generated in step S3. The priorities are divided into three levels: high, medium, and low. The mapping weight corresponding to the high priority is 1.5, the medium priority is 1.0, and the low priority is 0.5. The calculation method of the initial compensation coefficient is the real-time deviation rate multiplied by the mapping weight. For example, if the real-time deviation rate is 6.25% and the conflict priority is high, the initial compensation coefficient is 6.25% × 1.5 = 9.375%. If the initial compensation coefficient exceeds the preset upper limit value of 30%, it is forcibly truncated to 30%.

[0036] Adjust the initial compensation coefficient through the protocol type adaptive function. The protocol type adaptive function scales the initial compensation coefficient according to the network protocol type of the dynamic access node identifier. Under the Transmission Control Protocol, the scaling factor is the reciprocal of the path redundancy coefficient. For example, if the path redundancy coefficient is 0.6, the scaling factor is 1 ÷ 0.6 ≈ 1.6667; under the User Datagram Protocol, the scaling factor is the square root of the node connectivity. For example, if the node connectivity is 3, the scaling factor is 1.732. The scaled initial compensation coefficient is the original coefficient multiplied by the scaling factor. For example, if the initial compensation coefficient is 9.375% and the protocol type is the Transmission Control Protocol, the adjusted coefficient is 9.375% × 1.6667 ≈ 15.625%. If the scaled coefficient exceeds 30% or is less than 0, it is forcibly set to 30% or 0 through the amplitude limiting function.

[0037] Integrate the adjusted initial compensation coefficient with the network topology toughness input deviation fusion model to generate the target deviation compensation value. The deviation fusion model constrains the effective range of the target deviation compensation value through the product of the path redundancy coefficient and the node connectivity to ensure that the compensation value does not exceed the preset maximum deviation threshold; for example, if the path redundancy coefficient is 0.6 and the node connectivity is 3, the product is 0.6 × 3 = 1.8, and the effective range of the target deviation compensation value is the original compensation value multiplied by 1.8. If the adjusted initial compensation coefficient is 15.625%, the target deviation compensation value is 15.625% × 1.8 = 28.125%. The preset maximum deviation threshold is 30%. If the calculation result exceeds the threshold, the target deviation compensation value is forcibly set to 30%. The calculation result is rounded to two decimal places and written into the control instruction queue for the electronic thermostat valve opening adjustment module to call.

[0038] The co - design of S3 and S4 solves the coupling problem of control instruction conflicts and temperature control deviation in a dynamic network environment. Traditional methods rely on static network parameters or single metrics (such as latency, location) for conflict prediction and compensation, and it is difficult to adapt to the dynamic changes of the vehicle - to - everything (V2X) network topology and the real - time requirements of hydrogen fuel engine thermal management. S3 quantifies network stability through dynamic synchronization coefficients, combines causal influence to analyze implicit associations between nodes, and accurately identifies spatio - temporal conflict regions; S4 evaluates the network's anti - interference ability based on network topology resilience (path redundancy and connectivity), combines real - time temperature deviation to generate compensation values, and ensures that the temperature control instructions can still maintain thermal balance when the link fluctuates. Compared with the existing technology, through dynamic parameter fusion and multi - dimensional network resilience analysis, it improves the accuracy of conflict prediction and reduces the temperature control response delay, solves the efficiency loss problem caused by the disconnection between network status and thermal management, deeply couples the network dynamic characteristics with physical control parameters, and breaks through the limitations of traditional hierarchical design.

[0039] S5. Gradient - correct the total heating success rate instruction according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat. The specific implementation is as follows: Generate a gradient correction step size based on the current change rate and historical change trend of the target deviation compensation value. The current change rate is the ratio of the absolute value of the difference of the target deviation compensation value within a preset time window to the time interval. The duration of the preset time window is dynamically configured according to the control period of the vehicle node, with a default value of 5 seconds. The time interval is the difference between the current moment and the previous calculation moment. For example, if the previous target deviation compensation value is 15% and the current value is 20%, and the time interval is 5 seconds, then the current change rate is |20% - 15%|÷5 = 1% / second. The historical change trend calculates the change direction of the target deviation compensation value in the past 3 time windows through the exponential moving average method. The decay coefficient of the exponential moving average method is dynamically adjusted according to the network delay, with a default value of 0.9. If the moving average value is positive, it is determined as an upward trend; if it is negative, it is determined as a downward trend. The generation method of the gradient correction step size is the current change rate multiplied by the trend weight coefficient. The trend weight coefficient is optimized and determined through experimental data. The upward trend weight coefficient is 1.2, and the downward trend is 0.8.

[0040] Adjust the correction amplitude according to the deviation direction between the gradient-corrected total heating success rate command and the current one. The deviation direction is determined by the positive or negative sign of the target deviation compensation value. If the target deviation compensation value is positive, the heating power needs to be increased; if it is negative, the power needs to be decreased. The adjustment method of the correction amplitude is the gradient correction step size multiplied by the direction coefficient. The direction coefficient is dynamically configured according to the engine heat load state. The positive direction coefficient is 1.5, and the negative direction coefficient is 0.5. For example, when the engine is in a high-load state, the positive direction coefficient is increased to 1.8 to accelerate the response. If the correction amplitude exceeds the preset single-step correction upper limit, which is set according to the maximum heating power limit of the electronic thermostat. For example, if a certain type of thermostat allows a maximum single adjustment amplitude of 2%, the correction amplitude is forcibly truncated to 2%.

[0041] Dynamically match the corrected total heating success rate command with the real-time temperature feedback value of the electronic thermostat. The real-time temperature feedback value is the coolant outlet temperature currently collected by the electronic thermostat. If the deviation between the predicted temperature corresponding to the corrected command and the real-time temperature exceeds the preset temperature tolerance, which is set to 5°C according to the engine model, then trigger dynamic matching correction. The correction method is to adjust the heating power command in the reverse proportion according to the deviation. For example, if the predicted temperature is 5°C higher than the real-time temperature, the heating power command is reduced by 1%. If the corrected command exceeds the preset heating power range, which is set to 0 - 100% according to the physical parameters of the electronic thermostat, then it is forcibly truncated to the boundary value through the limit function.

[0042] Reverse-derive the adjustment amount of the reference parameter of the electronic thermostat valve opening according to the corrected total heating success rate command. The adjustment amount of the reference parameter is the difference between the corrected command and the original command multiplied by the preset conversion coefficient. The preset conversion coefficient is calibrated according to the thermodynamic characteristics of the engine and the valve mechanical response speed. For example, the conversion coefficient of a certain type of hydrogen fuel engine is 0.3, indicating that every 1% change in heating power corresponds to a 0.3 percentage point change in valve opening. The adjustment amount is rounded to one decimal place and written into the temporary buffer. The buffer capacity is set to 10 commands according to the throughput capacity of the control command queue.

[0043] Eliminate the instantaneous jitter of the adjustment amount through the hysteresis compensation function of the valve opening controller. The hysteresis compensation function uses the moving average method to smooth the last 3 adjustment amounts. The window size of the moving average method is set to 3 times according to the mechanical inertia of the valve. For example, if the last 3 adjustment amounts are 0.6%, 0.5%, and 0.7% respectively, the smoothed adjustment amount is (0.6 + 0.5 + 0.7) ÷ 3 = 0.6%. The updated valve opening reference parameter is sent to the actuator through the control command queue. The scheduling strategy of the control command queue is optimized to be first-in-first-out according to the response delay of the actuator to ensure that the instruction execution order is consistent with the generation order.

[0044] S6. The distributed control unit of the Internet of Things node issues the corrected total heating power success rate instruction and the valve opening reference parameter, and updates the communication reliability weight. The specific implementation is as follows: Generate a multicast transmission strategy based on the priority levels of the instructions in the conflict priority sorting table and the network topology resilience parameters. The conflict priority sorting table is generated in step S3. The priority levels are divided into three levels: high, medium, and low. The network topology resilience parameters include the path redundancy coefficient and the node connectivity. The path redundancy coefficient is the ratio of the number of available communication paths to the total number of paths, and the node connectivity is the number of nodes that directly communicate with the dynamically accessed node. The multicast transmission strategy is generated by allocating high-priority instructions to the combined paths with high path redundancy coefficient and high node connectivity. For example, if the priority level of an instruction is high, the path redundancy coefficient is 0.8, and the node connectivity is 4, then the path with a redundancy coefficient greater than 0.7 and a connectivity greater than 3 is preferentially selected for distribution. Medium- and low-priority instructions dynamically select the remaining paths according to the real-time network load. If the network load exceeds 70%, the path degradation mechanism is triggered, and some low-priority instructions are routed to the backup path. The minimum redundancy coefficient requirement for the backup path is 0.4.

[0045] Select the instruction encoding format according to the network protocol type and the real-time load status. The segmented acknowledgment retransmission mechanism is adopted under the Transmission Control Protocol. The segmented acknowledgment retransmission mechanism splits the instruction into data segments of a fixed length, attaches a unique sequence number to each segment, deletes the cache at the sender after the receiver acknowledges in order, and retransmits the corresponding data segment if the acknowledgment is not received within the timeout period. For example, the length of each segment is 1024 bytes, the acknowledgment timeout is 500 milliseconds, and the sequence number increases in the sending order. The redundant data packet broadcast mechanism is adopted under the User Datagram Protocol. The redundant data packet broadcast mechanism sends the key instruction packet three times redundantly. The receiver parses the content according to the first correctly received packet, and the subsequent duplicate packets are automatically discarded. For example, the total heating power success rate instruction packet is sent continuously three times, with an interval of 50 milliseconds. The receiver processes immediately after receiving the first valid packet, and the subsequent duplicate packets are only used for fault tolerance.

[0046] Schedule the instruction distribution order through the queue management module of the distributed control unit. The scheduling rule is that high-priority instructions in the conflict priority sorting table are inserted into the head of the queue and allocated resources not less than 50% of the total bandwidth, medium-priority instructions are allocated 30%, and low-priority instructions are allocated 20%. If network congestion causes insufficient bandwidth, low-priority instructions are delayed until the next cycle for sending, and the delay time is dynamically adjusted according to the real-time load. For example, when the load is 80%, the delay time is 2 seconds, and when the load is 90%, the delay time increases to 5 seconds. The bandwidth allocation strategy of the queue management module is implemented through the token bucket algorithm. The token generation rate is positively correlated with the instruction priority. The token generation rate for high-priority instructions is 10 per second, for medium-priority instructions is 6 per second, and for low-priority instructions is 4 per second. When the tokens are exhausted, the instruction distribution is suspended.

[0047] Update the communication reliability weight after receiving the instruction confirmation signal from the execution mechanism. The update logic is to dynamically adjust the weight gain coefficient based on the confirmation signal response time and packet loss rate. When the response time is less than 200 milliseconds, the gain coefficient is 1.2; when it is between 200 and 500 milliseconds, it is 1.0; when it is greater than 500 milliseconds, it is 0.8. When the packet loss rate is below 5%, the attenuation amplitude of the gain coefficient is 0.1; when it is between 5% and 10%, it is 0.3; when it is above 10%, it is 0.5. For example, if the response time of a certain node is 150 milliseconds and the packet loss rate is 3%, then the communication reliability weight is updated to the original value multiplied by 1.2 and then subtracted by 0.1. The updated weight value is rounded to the nearest integer and restricted within the range of 0 to 100. If the calculation result exceeds 100, it is forced to be set to 100; if it is less than 0, it is set to zero.

[0048] Synchronize the updated communication reliability weight to the node registry of the cloud server. The field definitions of the node registry include the vehicle unique code, network address, communication reliability weight, and path redundancy coefficient, which are exactly the same as the results of the network topology resilience analysis in step S4. The synchronization process is completed through the encrypted channel of the Secure Sockets Layer protocol. The encryption algorithm uses AES-256, and the key exchange protocol is ECDHE-RSA. The updated weight data is broadcast to all associated nodes within 500 milliseconds. The heartbeat packet sending interval is dynamically adjusted according to the network stability. The interval is 5 seconds under a stable network and shortened to 1 second under a fluctuating network to ensure the real-time consistency of the topology data. If no response is received for three consecutive heartbeat packets, the node is determined to be offline and the topology status is updated.

[0049] This embodiment solves the problem of the disconnection between the network state and the thermal management logic in the traditional temperature control system of hydrogen fuel engines through the deep coupling and real-time feedback mechanism of multi-dimensional dynamic parameters. Traditional methods rely on static network configurations or single metrics (such as latency, temperature) for control, making it difficult to adapt to the real-time requirements of the dynamic changes in the vehicle network topology and complex working conditions. This embodiment uses the dynamic synchronization coefficient as a bridge between network stability and thermal control instructions, combines causal influence analysis to identify hidden associations between nodes, and accurately predicts spatio-temporal conflicts. At the same time, it introduces the network topology resilience to evaluate the dynamic anti-interference ability of path redundancy and node connectivity, ensuring that the compensation instructions can still respond accurately when the link fluctuates. The weight fusion and gradient correction mechanism with adaptive protocol types further deeply binds the network state (such as TCP / UDP differences) and physical control parameters (heating power, valve opening), breaking through the response delay bottleneck of the traditional hierarchical design. The synergistic effect of each step realizes the synchronous optimization of network load balancing, temperature control stability, and hardware life, and constructs an efficient thermal management architecture adapted to dynamic heterogeneous networks through the multi-dimensional fusion and real-time feedback closed-loop of dynamic parameters.

[0050] All the calculations involved in the embodiment are dimensionless numerical calculations, and the preset parameters and threshold selections in the calculations are set by those skilled in the art according to the actual situation.

[0051] It should be noted that the present invention can be deployed on the device itself to achieve embedded applications, or can also run on a PC or other terminal with a user interface, so as to meet various hardware environments and usage requirements.

[0052] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0053] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0054] In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or modules can be in electrical, mechanical or other forms.

[0055] The module described as a separation component may or may not be physically separated. The component shown as a module may or may not be a physical module, and it may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0056] In addition, in each embodiment of this application, the various functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0057] If the above functions are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0058] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all of them should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0059] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An electronic thermostat heating assembly control method for a hydrogen fuel engine based on the Internet of Things, characterized in that, Including: S1. Real-time collect the temperature data of the electronic thermostat of the hydrogen fuel engine and the network status data of the Internet of Things nodes. The network status data includes the dynamically accessed node identifier and the data packet transmission delay; S2. Generate the dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamically accessed node identifier and the data packet transmission delay; S3. Based on the dynamic synchronization coefficient and the causal influence in the network topology corresponding to the dynamically accessed node identifier, predict the spatio-temporal conflict area of the multi-node control instruction and divide the conflict priority; S4. When the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier, and generate the target deviation compensation value by combining the electronic thermostat temperature data and the conflict priority; S5. Gradient-correct the total heating success rate instruction according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat; S6. Send the corrected total heating success rate instruction and the valve opening reference parameter through the distributed control unit of the Internet of Things node, and update the communication reliability weight.

2. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 1, characterized in that, Real-time collect the temperature data of the electronic thermostat of the hydrogen fuel engine and the network status data of the Internet of Things nodes, including: Obtain the electronic thermostat temperature data through the vehicle-mounted communication terminal; Based on the vehicle kinematic parameters corresponding to the dynamically accessed node identifier, real-time analyze the network status data of the Internet of Things nodes; Calculate the data packet transmission delay by the time stamp difference method; Store the electronic thermostat temperature data, the dynamically accessed node identifier and the data packet transmission delay into the distributed cache queue.

3. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 1, wherein, Generate the dynamic synchronization coefficient of the current control link according to the communication reliability weight of the dynamically accessed node identifier and the data packet transmission delay, including: Normalize the communication reliability weight of the dynamically accessed node identifier; Generate the dynamic delay correction factor according to the stability evaluation result of the data packet transmission delay. The stability evaluation is realized by calculating the ratio of the variance to the mean value of the data packet transmission delays of a continuous preset number; Input the normalized communication reliability weight and the dynamic delay correction factor into the weighted fusion function to generate the dynamic synchronization coefficient. The weighted fusion function dynamically adjusts the fusion ratio of the normalized communication reliability weight and the dynamic delay correction factor based on the current network protocol type; Update the communication reliability weight according to the real-time change rate of the dynamic synchronization coefficient.

4. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 1, characterized in that, Based on the dynamic synchronization coefficient and the causal influence in the network topology corresponding to the dynamically accessed node identifier, predict the spatio-temporal conflict area of the multi-node control instruction and divide the conflict priority, including: Extract the real-time stability parameter of the control link corresponding to the dynamic synchronization coefficient; Construct the causal influence graph according to the communication path dependence of the dynamically accessed node identifier; Analyze the spatio-temporal attribute distribution of the node instructions in the causal influence graph through the spatio-temporal overlap detection algorithm to identify the spatio-temporal conflict area; Generate the conflict priority ranking table based on the real-time stability parameter and the historical conflict probability of the spatio-temporal conflict area. The sorting rule of the conflict priority ranking table is that the area where the real-time stability parameter is lower than the preset stability threshold and the historical conflict probability is higher than the preset probability threshold is assigned the highest conflict priority; Dynamically match the conflict priority sorting table with the frequency response characteristics of the dynamic synchronization coefficient, and adjust the priority division threshold according to the periodic and burst characteristics of the control instructions.

5. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 4, characterized in that, The real-time stability parameter is generated through the collaborative analysis of the historical volatility of the dynamic synchronization coefficient and the current network load status. The historical volatility is the standard deviation of the dynamic synchronization coefficient within a preset past time period, and the current network load status is the ratio of the number of data packets received by the vehicle node per unit time to the total bandwidth.

6. The method for controlling a hydrogen fuel engine electronic thermostat heating assembly based on the Internet of Things according to claim 4 is characterized in that: The communication path dependence is generated by the non-linear combination of the covariance matrix of the historical communication success rate between nodes and the path redundancy coefficient. The path redundancy coefficient is the ratio of the number of available communication paths between nodes to the total number of paths.

7. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 1, characterized in that When the dynamic synchronization coefficient is lower than the preset synchronization threshold, analyze the network topology resilience of the dynamically accessed node identifier, and generate a target deviation compensation value in combination with the electronic thermostat temperature data and the conflict priority, including: When the dynamic synchronization coefficient is lower than the preset synchronization threshold, parse the network topology resilience of the dynamically accessed node identifier. The network topology resilience is generated through the non-linear association of the path redundancy coefficient and the node connectivity. Generate an initial compensation coefficient according to the mapping relationship between the real-time deviation rate of the electronic thermostat temperature data and the conflict priority. Adjust the initial compensation coefficient through the protocol type adaptive function. The protocol type adaptive function scales the initial compensation coefficient according to the network protocol type of the dynamically accessed node identifier. The scaling factor is the reciprocal of the path redundancy coefficient under the Transmission Control Protocol. Input the adjusted initial compensation coefficient and the network topology resilience into the deviation fusion model to generate the target deviation compensation value.

8. The method for controlling a hydrogen fuel engine electronic thermostat heating assembly based on the Internet of Things according to claim 1, characterized in that: Perform gradient correction on the total heating success rate instruction according to the target deviation compensation value, and synchronously update the valve opening reference parameter of the electronic thermostat, including: Generate a gradient correction step size based on the current change rate and the historical change trend of the target deviation compensation value. Adjust the correction amplitude according to the gradient correction step size and the deviation direction of the current total heating success rate instruction. The deviation direction is determined by the positive and negative signs of the target deviation compensation value. Dynamically match the corrected total heating success rate instruction with the real-time temperature feedback value of the electronic thermostat. If the corrected instruction exceeds the preset heating power range, it is forced to be truncated to the boundary value through the clipping function. Reverse-derive the adjustment amount of the valve opening reference parameter of the electronic thermostat according to the corrected total heating success rate instruction. Eliminate the instantaneous jitter of the adjustment amount through the hysteresis compensation function of the valve opening controller, and write the updated valve opening reference parameter into the control instruction queue and send it to the actuator.

9. The control method of the electronic thermostat heating assembly of a hydrogen fuel engine based on the Internet of Things according to claim 1, wherein, Send the corrected total heating success rate instruction and the valve opening reference parameter through the distributed control unit of the Internet of Things node, and update the communication reliability weight, including: Generate a multicast transmission strategy based on the priority levels of each instruction in the conflict priority sorting table and the network topology resilience parameters. Select the instruction encoding format according to the network protocol type and the real-time load status. The segment acknowledgment retransmission mechanism is adopted under the Transmission Control Protocol, and the redundant data packet broadcast mechanism is adopted under the User Datagram Protocol. The queue management module of the distributed control unit schedules the order of issuing instructions. The scheduling rule is that high-priority instructions in the conflict priority sorting table are given priority to allocate high-bandwidth resources, and low-priority instructions are sent later. After receiving the command confirmation signal from the actuator, the communication reliability weight is updated. The update logic is to adjust the weight gain coefficient based on the dynamic feedback of the confirmation signal response time and packet loss rate; The updated communication reliability weight is synchronized to the node registry of the cloud server and broadcast to the associated nodes through the heartbeat packet mechanism to maintain topology consistency.

Images