A multi-source signal fusion thermal management system comprehensive control method and system

Abstract

The present application relates to the field of new energy vehicle thermal management technology, and belongs to a kind of multi-source signal fusion thermal management system comprehensive control method and system.The method is unified access to vehicle CAN, LIN, hard-wired sensor and actuator feedback and other multi-source signals through multi-protocol communication interface, and establishes a unified signal database;After pre-processing and time synchronization of multi-source signals, redundant signal groups are automatically identified;Real-time quality assessment is performed on each signal source, and a comprehensive health score is calculated;An adaptive fusion algorithm based on health square weighting is used to generate a high-reliability fusion estimate value;When a single sensor fails, the weight of the faulty signal is automatically reduced or removed, and the remaining healthy signal sources are used to continue weighted fusion calculation.The present application significantly improves the reliability of the thermal management system signal, and realizes accurate estimation and continuous and stable operation of the system when the sensor fails.

Description

Technical Field

[0001] This invention relates to the field of thermal management technology for new energy vehicles, and belongs to a comprehensive control method and system for a thermal management system that integrates multi-source signals. Background Technology

[0002] With the rapid development of new energy vehicle technology, the complexity of vehicle thermal management systems (TMS) is increasing. Modern thermal management systems need to simultaneously manage multiple subsystems such as battery thermal management, electric drive system cooling, cabin air conditioning, and charging interface temperature control, involving a large number of sensor signals and actuator control.

[0003] In existing technologies, the signal acquisition of thermal management systems mainly relies on a single sensor or a single communication bus, which has the following technical drawbacks: First, insufficient signal reliability: When a single sensor fails, the system cannot obtain accurate physical quantity measurements, which may lead to incorrect control decisions and thus cause safety hazards such as thermal runaway.

[0004] Second, the signal sources are scattered: the vehicle CAN bus, the thermal management dedicated CAN bus (TMS_CAN), the LIN bus, and the feedback signals from hard-wired sensors and actuators are scattered in different communication links, lacking a unified management and fusion mechanism.

[0005] Third, the lack of redundancy design: Traditional schemes usually use simple signal selection logic (such as taking the maximum or minimum value), which fails to make full use of the correlation and complementarity between multiple source signals to make the optimal estimate.

[0006] Fourth, fault diagnosis is delayed: existing solutions often only trigger fault diagnosis after the sensor has completely failed, making it impossible to identify and take compensatory measures in advance during the sensor performance degradation stage.

[0007] Therefore, there is an urgent need for a comprehensive control method for thermal management systems that can uniformly manage multi-source heterogeneous signals, improve signal reliability through intelligent fusion algorithms, and provide accurate estimates when sensors fail. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a comprehensive control method and system for a thermal management system that integrates multi-source signals, solving technical issues in the prior art such as insufficient signal reliability, lack of unified management of multi-source signals, and inability to provide accurate estimates when sensors fail.

[0009] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a comprehensive control method and system for a thermal management system that integrates multi-source signal fusion, comprising the following steps: Step S1: The thermal management system controller accesses multiple source signals from the vehicle through a multi-protocol communication interface; Step S2: Preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. Step S3: Based on the signal semantic description and physical quantity mapping relationship stored in the thermal management system controller, automatically identify signal groups with redundant relationships, and classify signals that measure the same physical quantity or have strong correlation into the same redundant signal group; Step S4: Perform real-time quality assessment on each signal source within each redundant signal group, calculate the health index of each signal, and sum them by weight to obtain the overall health score; Step S5: Based on the comprehensive health status, an adaptive weighted fusion algorithm is used to fuse the signals in the redundant signal group to generate a fusion estimate; wherein, the fusion weight of the adaptive weighted fusion algorithm is dynamically adjusted according to the real-time health status of each signal, and the dynamic adjustment includes weight smoothing adjustment caused by gradual changes in health status and fault detection and fault tolerance processing; Step S6: Based on the fused estimated value, execute the thermal management system control strategy, generate actuator control commands, and send them to each actuator through the corresponding communication interface.

[0010] Preferably, in step S2, the signal preprocessing includes the following sub-steps: Filtering and noise reduction: High-frequency noise is removed from analog sensor signals by using moving average filtering or Kalman filtering; reasonable range checks are performed on digital communication signals to eliminate obvious outliers; Dimension conversion: Converting the same physical quantity signal from different sources into standard engineering units; for example, the coolant temperature signal from the CAN bus (represented by the raw CAN value) and the voltage signal collected by the NTC sensor are both converted to degrees Celsius (°C). Time synchronization: Based on the internal clock of the thermal management system controller, the signals of each communication bus are timestamped and the time synchronization accuracy is controlled within ±5ms.

[0011] Preferably, the time synchronization, specifically including the alignment of signal inter-stamps for each communication bus, includes: For CAN / LIN signals, the timestamps of the bus messages are used for backtracking alignment. For hard-wired sensor signals, interpolation alignment is performed based on the sampling period.

[0012] Preferably, in step S3, the identification of redundant signals is based on the following principles: Principle 1: Direct Redundancy: Multiple sensors or signal sources directly measure the same physical quantity; for example, the battery pack inlet coolant temperature can be directly measured by the NTC sensor of the thermal management system controller, and can also be provided by the battery management system (BMS) via the CAN bus.

[0013] Principle 2: Indirect Redundancy: There are known physical or empirical relationships between different physical quantities, and redundancy estimates can be obtained through model calculations; for example, the compressor outlet pressure can be indirectly calculated through parameters such as compressor speed, ambient temperature, and condenser air volume.

[0014] Principle 3: Spatial redundancy: There is a strong correlation between multiple spatially adjacent measurement points of the same physical quantity; for example, the measurement values ​​between adjacent temperature sensors in a battery module have limited differences under normal operating conditions.

[0015] Preferably, in step S4, the quality assessment of each signal source includes the following health indicators: Signal rate of change reasonableness: Calculate the signal derivative and check whether it exceeds the physical maximum allowable rate of change. Within the reasonable range, the health score is 1.0. If it exceeds the range, it will be linearly decayed according to the degree of deviation. Consistency of signals from the same source: Calculate the deviation between the signal and other signal sources in the same group, take the normalized consistency index, and the health score is 1.0 when the deviation is less than the threshold, and linearly decreases as the deviation increases when it exceeds the threshold; Historical statistical characteristics: Compare the current value with the mean and variance of the historical sliding window to detect abnormal deviations. Within ±3σ, the health score is 1.0; outside the range, it is reduced by the deviation σ. Communication link status: Monitor the loss rate, error frame rate and timeout status of CAN / LIN messages. The health status is 1.0 when there are no errors and decreases linearly to 0 when the error rate increases. Signal validity flag: The validity / fault flag bit reported by the signal source device itself is taken. When the validity flag is 1, the health level is 1.0, and when it is invalid or faulty, the health level is 0. The overall health score is obtained by weighting and summing the above indicators. The weights are configured based on the signal type and operating conditions.

[0016] Preferably, in step S5, the weight smoothing adjustment rule caused by the gradual change in health is as follows: The fusion estimate is: ; in, These are the measured values ​​from each signal source; For weighting; For the current weight to be calculated One sensor; This is an index variable for all sensors within the same redundant group, and iterates through all n sensors in that group; n is the total number of sensors in that redundant group. Fusion weights: ; in, This refers to the overall health of the sensor.

[0017] Preferably, the adaptive weighted fusion algorithm also introduces a confidence interval verification mechanism: if the measured value of a certain signal source is... The deviation from the fusion estimate exceeds a preset threshold (configurable). If the signal is within a fixed engineering range, it will be temporarily blocked and will not participate in the fusion calculation of the current cycle.

[0018] Preferably, in step S5, the fault detection includes a three-level mechanism: Level 1, Performance Degradation Warning: When the overall health of the signal source... If the signal source is in the range of 0.5 to 0.8 for N consecutive periods (default N=10, i.e., 100ms), it is determined that the signal source is in a state of performance degradation. The system issues a warning message and automatically reduces the weight of the signal, suggesting predictive maintenance, but the signal still participates in the fusion calculation.

[0019] Level 2, Fault Confirmation and Isolation: When the overall health of the signal source... If the value is below 0.3 for M consecutive cycles (default M=5, i.e., 50ms), the signal source is determined to be faulty. The system isolates the signal source from the fusion calculation, resets the weight to zero, and only uses the remaining healthy signal sources for fusion estimation. At the same time, the fault code is recorded and reported to the vehicle diagnostic system. Level 3, Insufficient Redundancy Alarm: When the number of healthy signal sources in the redundant signal group drops to less than 1, the system issues an insufficient redundancy alarm. At this time, the system will switch to the safety degradation mode and use the last effective fusion value combined with the physical model prediction value for conservative control.

[0020] Preferably, in the safety degradation mode, the control strategy switches to conservative mode, limiting the maximum output capability of the actuator to prevent over-control due to signal estimation uncertainty.

[0021] Preferably, in step S6, the thermal management system control strategy is executed based on the fused estimate, specifically including: Control target value calculation: Based on the fused estimated values, the control target values ​​of each actuator are calculated using the PID control algorithm, including compressor speed, water pump flow rate, fan speed, electronic expansion valve opening degree, and heater power; Control command generation and issuance: The calculated control target values ​​are converted into control commands that match the communication protocol of each actuator, and then issued to the local controller of each actuator through the corresponding communication interface; Execution feedback acquisition: Collect the actual operating status parameters of each actuator, compare them with the corresponding target values, and calculate the control deviation of each actuator; Control strategy optimization: The actual operating status parameters of each actuator, control deviation, and signal source health information are used as feedback data and sent back for signal fusion weight adjustment and control algorithm parameter correction in the next control cycle.

[0022] On the other hand, the present invention also provides a comprehensive control system for a thermal management system that integrates multi-source signal fusion to implement the above method, specifically including: The multi-protocol communication interface layer is responsible for unified access to all communication links and hardwired signals of the vehicle. The signal preprocessing layer is used to preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. The signal grouping management layer automatically identifies signal groups with redundant relationships based on signal semantic descriptions and physical quantity mapping relationships, and groups signals that measure the same physical quantity or have strong correlations into the same redundant signal group. The fusion computing layer receives multi-source signals after processing by the signal grouping management layer, performs real-time quality assessment and weighted fusion calculation on each signal source, and generates a fusion estimate. The control decision layer executes the thermal management system control strategy based on the fused signal estimates, including boundary hysteresis verification and anti-jitter processing for mode switching, generates actuator control commands, and sends them to each actuator through the corresponding communication interface.

[0023] Compared with the prior art, the present invention provides a comprehensive control method and system for a thermal management system that integrates multi-source signal fusion, which has the following beneficial effects: This invention unifies the access of multiple signal sources, including CAN, LIN, hard-wired sensors, and actuator feedback from the entire vehicle, through a multi-protocol communication interface, establishing a unified signal database. After preprocessing and time synchronization of the multi-source signals, redundant signal groups are automatically identified. Real-time quality assessment is performed on each signal source, calculating a comprehensive health score. An adaptive fusion algorithm based on squared health weighting is used to generate a highly reliable fusion estimate. When a single sensor fails, the weight of the faulty signal is automatically reduced or removed, and the remaining healthy signal sources are used to continue the weighted fusion calculation. This invention significantly improves the signal reliability of the thermal management system, achieving accurate estimation in the event of sensor failure and ensuring continuous and stable system operation.

[0024] The features and advantages of the present invention will be described in detail through embodiments and in conjunction with the accompanying drawings. Attached Figure Description

[0025] Figure 1 This invention provides a comprehensive control method for a thermal management system based on multi-source signal fusion, and a flowchart of the system. Figure 2 This is a flowchart of the unified access and preprocessing of multiple signals according to the present invention; Figure 3 This is a flowchart of the signal quality assessment process of the present invention; Figure 4 A schematic diagram showing the comparison of fusion estimation before and after a single sensor failure; Figure 5 This is a schematic diagram of the system architecture of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0027] See Figure 1 This embodiment provides a comprehensive control method and system for a thermal management system that integrates multi-source signal fusion, including the following steps: Step S1: The thermal management system controller accesses multiple source signals from the vehicle through a multi-protocol communication interface; Step S2: Preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. Step S3: Based on the signal semantic description and physical quantity mapping relationship stored in the thermal management system controller, automatically identify signal groups with redundant relationships, and classify signals that measure the same physical quantity or have strong correlation into the same redundant signal group; Step S4: Perform real-time quality assessment on each signal source within each redundant signal group, calculate the health index of each signal, and sum them by weight to obtain the overall health score; Step S5: Based on the comprehensive health status, an adaptive weighted fusion algorithm is used to fuse the signals in the redundant signal group to generate a fusion estimate; wherein, the fusion weight of the adaptive weighted fusion algorithm is dynamically adjusted according to the real-time health status of each signal, and the dynamic adjustment includes weight smoothing adjustment caused by gradual changes in health status and fault detection and fault tolerance processing; Step S6: Based on the fused estimated value, execute the thermal management system control strategy, generate actuator control commands, and send them to each actuator through the corresponding communication interface.

[0028] In this embodiment, the thermal management system controller is a TMS controller.

[0029] Specifically, see Figure 2 In step S2, signal preprocessing includes the following sub-steps: Filtering and noise reduction: High-frequency noise is removed from analog sensor signals by using moving average filtering or Kalman filtering; reasonable range checks are performed on digital communication signals to eliminate obvious outliers; Dimension conversion: Converting the same physical quantity signal from different sources into standard engineering units; for example, the coolant temperature signal from the CAN bus (represented by the raw CAN value) and the voltage signal collected by the NTC sensor are both converted to degrees Celsius (°C). Time synchronization: Based on the internal clock of the thermal management system controller, the signals of each communication bus are timestamped and the time synchronization accuracy is controlled within ±5ms.

[0030] Specifically, the time synchronization, including the alignment of signal inter-stamps for each communication bus, specifically includes: For CAN / LIN signals, the timestamps of the bus messages are used for backtracking alignment. For hard-wired sensor signals, interpolation alignment is performed based on the sampling period.

[0031] Specifically, in step S3, the identification of redundant signals is based on the following principles: Principle 1: Direct Redundancy: Multiple sensors or signal sources directly measure the same physical quantity; for example, the battery pack inlet coolant temperature can be directly measured by the NTC sensor of the thermal management system controller, and can also be provided by the battery management system (BMS) via the CAN bus.

[0032] Principle 2: Indirect Redundancy: There are known physical or empirical relationships between different physical quantities, and redundancy estimates can be obtained through model calculations; for example, the compressor outlet pressure can be indirectly calculated through parameters such as compressor speed, ambient temperature, and condenser air volume.

[0033] Principle 3: Spatial redundancy: There is a strong correlation between multiple spatially adjacent measurement points of the same physical quantity; for example, the measurement values ​​between adjacent temperature sensors in a battery module have limited differences under normal operating conditions.

[0034] Specifically, in step S4, the quality assessment of each signal source includes the following health indicators: Signal rate of change reasonableness: Calculate the signal derivative and check whether it exceeds the physical maximum allowable rate of change. Within the reasonable range, the health score is 1.0. If it exceeds the range, it will be linearly decayed according to the degree of deviation. Consistency of signals from the same source: Calculate the deviation between the signal and other signal sources in the same group, take the normalized consistency index, and the health score is 1.0 when the deviation is less than the threshold, and linearly decreases as the deviation increases when it exceeds the threshold; Historical statistical characteristics: Compare the current value with the mean and variance of the historical sliding window to detect abnormal deviations. Within ±3σ, the health score is 1.0; outside the range, it is reduced by the deviation σ. Communication link status: Monitor the loss rate, error frame rate and timeout status of CAN / LIN messages. The health status is 1.0 when there are no errors and decreases linearly to 0 when the error rate increases. Signal validity flag: The validity / fault flag bit reported by the signal source device itself is taken. When the validity flag is 1, the health level is 1.0, and when it is invalid or faulty, the health level is 0. The above indicators are weighted and summed to obtain the overall health score H(i), with the weights configured according to the signal type and operating conditions.

[0035] Specifically, see Figures 3-4 In step S5, the weight smoothing adjustment rule caused by the gradual change in health is as follows: The fusion estimate is: ; in, These are the measured values ​​from each signal source; For weighting; For the current weight to be calculated One sensor; This is an index variable for all sensors within the same redundant group, and iterates through all n sensors in that group; n is the total number of sensors in that redundant group. Fusion weights: ; in, This refers to the overall health of the sensor.

[0036] Specifically, the adaptive weighted fusion algorithm also introduces a confidence interval verification mechanism: if the measured value of a certain signal source is... The deviation from the fusion estimate exceeds a preset threshold (configurable). If the signal is within a fixed engineering range, it will be temporarily blocked and will not participate in the fusion calculation of the current cycle.

[0037] Specifically, in step S5, the fault detection includes a three-level mechanism: Level 1, Performance Degradation Warning: When the overall health of the signal source... If the signal source is in the range of 0.5 to 0.8 for N consecutive periods (default N=10, i.e., 100ms), it is determined that the signal source is in a state of performance degradation. The system issues a warning message and automatically reduces the weight of the signal, suggesting predictive maintenance, but the signal still participates in the fusion calculation.

[0038] Level 2, Fault Confirmation and Isolation: When the overall health of the signal source... If the value is below 0.3 for M consecutive cycles (default M=5, i.e., 50ms), the signal source is determined to be faulty. The system isolates the signal source from the fusion calculation, resets the weight to zero, and only uses the remaining healthy signal sources for fusion estimation. At the same time, the fault code is recorded and reported to the vehicle diagnostic system. Level 3, Insufficient Redundancy Alarm: When the number of healthy signal sources in the redundant signal group drops to less than 1, the system issues an insufficient redundancy alarm. At this time, the system will switch to the safety degradation mode and use the last effective fusion value combined with the physical model prediction value for conservative control.

[0039] Specifically, in the safety degradation mode, the control strategy switches to conservative mode, limiting the maximum output capability of the actuator to prevent over-control due to signal estimation uncertainty.

[0040] In step S6, based on the fused estimated value, the thermal management system control strategy is executed, specifically including: Control target value calculation: Based on the fused estimated values, the control target values ​​of each actuator are calculated using the PID control algorithm, including compressor speed, water pump flow rate, fan speed, electronic expansion valve opening degree, and heater power; Control command generation and issuance: The calculated control target values ​​are converted into control commands that match the communication protocol of each actuator, and then issued to the local controller of each actuator through the corresponding communication interface; Execution feedback acquisition: Collect the actual operating status parameters of each actuator, compare them with the corresponding target values, and calculate the control deviation of each actuator; Control strategy optimization: The actual operating status parameters of each actuator, control deviation, and signal source health information are used as feedback data and sent back for signal fusion weight adjustment and control algorithm parameter correction in the next control cycle.

[0041] See Figure 4 On the other hand, the present invention also provides a comprehensive control system for a thermal management system that integrates multi-source signal fusion to implement the above method, specifically including: The multi-protocol communication interface layer is responsible for unified access to all communication links and hardwired signals of the vehicle. The signal preprocessing layer is used to preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. The signal grouping management layer automatically identifies signal groups with redundant relationships based on signal semantic descriptions and physical quantity mapping relationships, and groups signals that measure the same physical quantity or have strong correlations into the same redundant signal group. The fusion computing layer receives multi-source signals after processing by the signal grouping management layer, performs real-time quality assessment and weighted fusion calculation on each signal source, and generates a fusion estimate. The control decision layer executes the thermal management system control strategy based on the fused signal estimates, including boundary hysteresis verification and anti-jitter processing for mode switching, generates actuator control commands, and sends them to each actuator through the corresponding communication interface.

[0042] The multi-protocol communication interface layer is responsible for the unified access of all communication links and hardwired signals in the vehicle, including: (a) Vehicle CAN bus: Accesses vehicle status information broadcast by the vehicle controller (VCU), including vehicle speed, accelerator pedal opening, braking status, ambient temperature, etc.

[0043] (b) Thermal Management Dedicated CAN Bus (TMS_CAN): Connects high-speed communication signals between various sub-controllers within the thermal management system, including compressor speed, electric water pump flow rate, and electronic expansion valve opening.

[0044] (c) LIN bus: Connects low-speed sensor and actuator signals, such as damper actuator position, sunlight sensor irradiance, etc.

[0045] (d) Hardwired sensors: Analog / digital sensors that are directly connected to the TMS controller, including NTC temperature sensors, pressure sensors, flow sensors, etc.

[0046] (e) Actuator feedback: The actual status feedback signals of various actuators, such as the actual speed of the compressor, the actual opening degree of the valve, and the actual flow rate of the water pump.

[0047] To further illustrate the effects of the present invention, specific application examples are as follows: Taking the battery pack inlet coolant temperature signal as an example, the specific application of the method of the present invention is illustrated below: Redundant signal sources include: Under normal operating conditions, the health of the three signal sources is 1.0, the fusion weight is 1 / 3 for each, and the fusion estimate is close to the average of the three.

[0048] When an open-circuit fault occurs in the NTC sensor at the TMS controller inlet, the signal health status rapidly drops to 0 (both the communication link status indicator and the signal validity flag are 0), and its fusion weight automatically drops to 0. The weights of the remaining two signal sources are each adjusted to 1 / 2, and the fusion estimate is based on a weighted average of the temperature reported by the BMS and the temperature estimated by the thermal model.

[0049] The test results show that after the NTC sensor completely fails, the deviation between the fused output temperature and the actual temperature is controlled within ±1.5℃, which meets the control accuracy requirements of the thermal management system (±2℃).

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions or improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A comprehensive control method for a thermal management system based on multi-source signal fusion, characterized in that: Includes the following steps: Step S1: The thermal management system controller accesses multiple source signals from the vehicle through a multi-protocol communication interface; Step S2: Preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. Step S3: Based on the signal semantic description and physical quantity mapping relationship stored in the thermal management system controller, automatically identify signal groups with redundant relationships, and classify signals that measure the same physical quantity or have strong correlation into the same redundant signal group; Step S4: Perform real-time quality assessment on each signal source within each redundant signal group, calculate the health index of each signal, and sum them by weight to obtain the overall health score; Step S5: Based on the comprehensive health status, an adaptive weighted fusion algorithm is used to fuse the signals in the redundant signal group to generate a fusion estimate; wherein, the fusion weight of the adaptive weighted fusion algorithm is dynamically adjusted according to the real-time health status of each signal, and the dynamic adjustment includes weight smoothing adjustment caused by gradual changes in health status and fault detection and fault tolerance processing; Step S6: Based on the fused estimated value, execute the thermal management system control strategy, generate actuator control commands, and send them to each actuator through the corresponding communication interface.

2. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 1, characterized in that: In step S2, signal preprocessing includes the following sub-steps: Filtering and noise reduction: High-frequency noise is removed from analog sensor signals by using moving average filtering or Kalman filtering; reasonable range checks are performed on digital communication signals to eliminate obvious outliers; Dimensional conversion: Converting the same physical quantity signal from different sources into standard engineering units; Timestamp alignment: Based on the internal clock of the thermal management system controller, the signals of each communication bus are timestamped, and the time synchronization accuracy is controlled within ±5ms.

3. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 2, characterized in that: The time-stamp alignment of signals on each communication bus specifically includes: For CAN / LIN signals, the timestamps of the bus messages are used for backtracking alignment. For hard-wired sensor signals, interpolation alignment is performed based on the sampling period.

4. The integrated control method for a multi-source signal fusion thermal management system according to claim 1, characterized in that: In step S3, the identification of redundant signals is based on the following principles: Principle 1: Direct Redundancy: Multiple sensors or signal sources directly measure the same physical quantity; Principle 2: Indirect Redundancy: There are known physical or empirical relationships between different physical quantities, and redundancy estimates can be obtained through model calculations; Principle 3: Spatial redundancy: There is a strong correlation between multiple spatially adjacent measurement points of the same physical quantity.

5. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 1, characterized in that: In step S4, the quality assessment of each signal source includes the following health indicators: Signal rate of change reasonableness: Calculate the signal derivative and check whether it exceeds the physical maximum allowable rate of change. Within the reasonable range, the health score is 1.

0. If it exceeds the range, it will be linearly decayed according to the degree of deviation. Consistency of signals from the same source: Calculate the deviation between the signal and other signal sources in the same group, take the normalized consistency index, and the health score is 1.0 when the deviation is less than the threshold, and linearly decreases as the deviation increases when it exceeds the threshold; Historical statistical features: By comparing the current value with the mean and variance of historical sliding windows, abnormal shifts are detected. Health is 1.0 within the specified range; outside the range, it is considered a deviation. Number decay, among which The standard deviation of the signal values ​​within the historical sliding window; Communication link status: Monitor the loss rate, error frame rate and timeout status of CAN / LIN messages. When there are no errors, the health status is 1.

0. When the error rate increases, it decreases linearly to 0. Signal validity flag: The validity / fault flag bit reported by the signal source device itself is taken. When the validity flag is 1, the health level is 1.0, and when it is invalid or faulty, the health level is 0. The overall health score is obtained by weighting and summing the above indicators. The weights are configured based on the signal type and operating conditions.

6. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 1, characterized in that: In step S5, the weight smoothing adjustment rule caused by the gradual change in health is as follows: Calculate the fusion estimate: ; in, These are the measured values ​​from each signal source; For weighting; For the current weight to be calculated One sensor; This is an index variable for all sensors within the same redundant group, and iterates through all n sensors in that group; n is the total number of sensors in that redundant group. Fusion weights: ; in, This refers to the overall health of the sensor.

7. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 6, characterized in that: The adaptive weighted fusion algorithm also introduces a confidence interval verification mechanism: if the measured value of a certain signal source... If the deviation from the fusion estimate exceeds a preset threshold, the signal will be temporarily blocked and will not participate in the fusion calculation for the current cycle.

8. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 1, characterized in that: In step S5, the fault detection and fault tolerance handling rules are as follows: Based on the comprehensive health status, fault detection and fault tolerance processing are performed on the signal sources. When a fault or performance degradation of a signal source is detected, its fusion weight is automatically reduced or it is removed. The remaining healthy signal sources are used to continue weighted fusion calculation to output accurate valuation results.

9. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 1, characterized in that: In step S5, the fault detection includes a three-level mechanism: Level 1, Performance Degradation Warning: When the overall health of the signal source... If the signal source is in the first threshold range for N consecutive periods, it is determined that the signal source is in a state of performance degradation. The system issues an early warning message and automatically reduces the weight of the signal, suggesting predictive maintenance, but the signal still participates in the fusion calculation. Level 2, Fault Confirmation and Isolation: When the overall health of the signal source... If the signal source is below the second threshold for M consecutive cycles, the system determines that the signal source has failed. The system isolates the signal source from the fusion calculation, resets the weight to zero, and only uses the remaining healthy signal sources for fusion estimation. At the same time, the fault code is recorded and reported to the vehicle diagnostic system. Level 3, Insufficient Redundancy Alarm: When the number of healthy signal sources in the redundant signal group drops below the third threshold, the system issues an insufficient redundancy alarm. At this time, the system will switch to the safety degradation mode and use the last effective fusion value combined with the physical model prediction value for conservative control.

10. The integrated control method for a thermal management system based on multi-source signal fusion according to claim 9, characterized in that: In the aforementioned safety degradation mode, the control strategy switches to a conservative mode, limiting the maximum output capability of the actuator to prevent over-control due to signal estimation uncertainty.

11. The integrated control method for a thermal management system based on multi-source signal fusion as described in claim 1, characterized in that: In step S6, based on the fused estimated value, the thermal management system control strategy is executed, specifically including: Control target value calculation: Based on the fused estimated values, the control target values ​​of each actuator are calculated using the PID control algorithm, including compressor speed, water pump flow rate, fan speed, electronic expansion valve opening degree, and heater power; Control command generation and issuance: The calculated control target values ​​are converted into control commands that match the communication protocol of each actuator, and then issued to the local controller of each actuator through the corresponding communication interface; Execution feedback acquisition: Collect the actual operating status parameters of each actuator, compare them with the corresponding target values, and calculate the control deviation of each actuator; Control strategy optimization: The actual operating status parameters of each actuator, control deviation, and signal source health information are used as feedback data and sent back for signal fusion weight adjustment and control algorithm parameter correction in the next control cycle.

12. The system of the integrated control method for a multi-source signal fusion thermal management system according to any one of claims 1-11, characterized in that: Specifically, it includes: The multi-protocol communication interface layer is responsible for unified access to all communication links and hardwired signals of the vehicle. The signal preprocessing layer is used to preprocess the incoming multi-source signals, including filtering and noise reduction, dimension conversion and timestamp alignment, to ensure that signals from different sources are fused and calculated on the same time reference. The signal grouping management layer automatically identifies signal groups with redundant relationships based on signal semantic descriptions and physical quantity mapping relationships, and groups signals that measure the same physical quantity or have strong correlations into the same redundant signal group. The fusion computing layer receives multi-source signals after processing by the signal grouping management layer, performs real-time quality assessment and weighted fusion calculation on each signal source, and generates a fusion estimate. The control decision layer executes the thermal management system control strategy based on the fused signal estimates, including boundary hysteresis verification and anti-jitter processing for mode switching, generates actuator control commands, and sends them to each actuator through the corresponding communication interface.

Images