Protection control system based on vehicle-mounted high-voltage electrical abnormal state recognition

Abstract

This invention relates to the field of electrical safety protection technology, specifically to a protection and control system based on vehicle high-voltage electrical abnormality identification, including the following modules: (1) Multi-dimensional data acquisition module, S1.1 sensor unit: including voltage sensor (measurement range 0~1000V, accuracy ±0.5%), current sensor (measurement range 0~500A, accuracy ±0.1%), temperature sensor (measurement range -40~150℃); This invention integrates multi-dimensional parameters to construct an abnormality identification model, which can identify hidden abnormalities such as insulation degradation and increased contact resistance 5~10s in advance, with a fault response time ≤100ms, avoiding the risk of fault expansion from the source, reducing the occurrence rate of vehicle high-voltage electrical faults by more than 80%, classifying abnormal states into three levels: slight, moderate, and severe, corresponding to the graded protection actions of "parameter adjustment - power reduction operation - emergency power cut-off", avoiding minor abnormalities triggering vehicle power cut-off, reducing the life loss of power battery due to frequent power cut-off by 30%, and significantly improving the driving experience.

Description

Technical Field

[0001] This invention relates to the field of electrical safety protection technology, specifically a protection and control system based on the identification of abnormal electrical conditions in vehicle-mounted high voltage systems. Background Technology

[0002] The high-voltage electrical system (voltage level 200-800V) of new energy vehicles is the core of the vehicle's power output and energy management, encompassing key components such as the power battery, motor controller, high-voltage distribution box, and charging system. With the increasing driving range of new energy vehicles and the widespread adoption of fast charging technology, the operating conditions of the high-voltage electrical system are becoming increasingly complex (high-current charging and discharging, high and low temperature environments, vibration and shock), posing a severe challenge to its safe operation.

[0003] Delayed anomaly identification and untimely fault response: Traditional high-voltage electrical protection systems mostly use threshold-triggered monitoring (such as overvoltage / overcurrent threshold alarms), which can only identify obvious faults that have already occurred. They cannot identify latent abnormal states such as insulation degradation and increased contact resistance in advance. The fault response delay is ≥500ms, which can easily lead to the expansion of the fault and cause fire and electric shock risks.

[0004] The protection strategy is simplistic and lacks hierarchical control: The existing system only has a single protection logic of "power off when there is a fault" and does not formulate a hierarchical protection strategy based on the severity of the abnormality. Even minor abnormalities (such as instantaneous overvoltage) can trigger the high voltage power-off of the whole vehicle, resulting in a decline in driving experience and vehicle breakdown. At the same time, frequent power-offs will damage the life of the power battery.

[0005] The identification dimension is limited and the misjudgment rate is high: It only monitors basic electrical parameters such as voltage and current, without combining multi-dimensional data such as temperature, vibration, and insulation impedance. It is easy to misjudge normal operating condition fluctuations (such as high current at the beginning of fast charging) as abnormalities, with a misjudgment rate of ≥8%, which increases ineffective protection actions and after-sales costs.

[0006] Without closed-loop control, the protection effect is uncontrollable: after abnormal state is identified, only power-off / power-reduction actions are performed. There is no real-time feedback and adjustment of the system state after the protection action, and it is impossible to confirm whether the protection measures are effective. This can easily lead to a situation where "the protection action is performed but the fault is not eliminated".

[0007] Poor adaptability and incompatibility with multiple vehicle models: The protection system is strongly coupled with the high-voltage architecture of the vehicle. Vehicles with different voltage levels (400V / 800V) and different topologies need to be developed separately, resulting in a long development cycle (≥6 months) and high adaptation costs.

[0008] Existing technologies have not yet formed an integrated protection system of "multi-dimensional real-time identification - hierarchical precise protection - closed-loop feedback control". Lagging anomaly identification and crude protection strategies have become the core bottlenecks restricting the safety of high-voltage electrical systems in new energy vehicles. To address this, a protection and control system based on the identification of abnormal states of on-board high-voltage electrical systems is proposed. Summary of the Invention

[0009] In view of this, the present invention provides a protection and control system based on the identification of abnormal conditions of vehicle-mounted high-voltage electrical systems, in order to solve or alleviate the technical problems existing in the prior art, and at least provide a beneficial option.

[0010] The technical solution of this invention is implemented as follows: a protection and control system based on vehicle-mounted high-voltage electrical abnormality identification, comprising the following modules:

[0011] (1) Multi-dimensional data acquisition module

[0012] S1.1 Sensor Unit: Includes voltage sensor (measurement range 0~1000V, accuracy ±0.5%), current sensor (measurement range 0~500A, accuracy ±0.1%), temperature sensor (measurement range -40~150℃, accuracy ±1℃), insulation resistance sensor (measurement range 0~1000MΩ, accuracy ±2%), and vibration sensor (measurement range 0~2000Hz, accuracy ±0.1g), which are deployed at key nodes such as the power battery output terminal, high voltage distribution box, motor controller, and charging interface, respectively.

[0013] S1.2 Data Preprocessing Unit: Filters (using Kalman filter algorithm), denoises, and normalizes the collected raw data, with a sampling frequency ≥10kHz to ensure data accuracy;

[0014] S1.3 Data transmission unit: The pre-processed data is transmitted to the anomaly identification module using a CANFD bus (transmission rate 8Mbps), with a transmission delay of ≤10ms;

[0015] (2) Abnormal state identification module

[0016] S2.1 Feature Extraction Unit: Extracts 15 types of feature parameters, including voltage fluctuation coefficient, current harmonic content, temperature change rate, insulation impedance attenuation rate, and vibration spectrum characteristics;

[0017] S2.2 Anomaly Identification Model Unit: An anomaly identification model is constructed based on the random forest algorithm. The model input consists of 15 types of feature parameters, and the output consists of the anomaly status level (slight / moderate / severe) and anomaly type (insulation degradation / overcurrent / overvoltage / poor contact, etc.).

[0018] S2.3 Real-time Judgment Unit: The identification result is updated every 10ms, and a warning is given 5-10 seconds in advance for latent anomalies (such as insulation resistance decreasing from 500MΩ to 200MΩ). The anomaly identification accuracy is ≥99%.

[0019] (3) Graded protection control module

[0020] S3.1 Grading Determination Unit: Preset Abnormal State Grading Threshold:

[0021] Minor abnormalities: A single parameter exceeds the normal range by less than 10%, such as voltage fluctuation ±5% or insulation resistance ≥200MΩ;

[0022] Moderate abnormality: Multiple parameters exceed the normal range by 10% to 30%, such as voltage fluctuation of ±10% and temperature increase of 5℃, and insulation resistance of 100~200MΩ;

[0023] Severe abnormality: Parameters exceed the normal range by more than 30% or dangerous operating conditions occur, such as overcurrent ≥150% of rated current, insulation resistance ≤100MΩ;

[0024] S3.2 Protection Strategy Execution Unit: Executes corresponding protection actions based on the anomaly level:

[0025] Minor anomalies: Adjust the power battery output voltage / current and cooling system power to bring the parameters back to the normal range;

[0026] Moderate abnormality: Control the motor controller to reduce power to 50% of rated power and limit fast charging current to 30% of rated current;

[0027] Severe anomaly: Triggers an emergency power outage of the high-voltage distribution box, cutting off the high-voltage circuit of the entire vehicle and activating the vehicle alarm system;

[0028] S3.3 Action Execution Drive Unit: Through hard-wired + CAN dual-channel control, it ensures a 100% success rate in executing protective actions with an execution delay of ≤50ms;

[0029] (4) Closed-loop feedback module

[0030] S4.1 Status Monitoring Unit: After the protective action is executed, key system parameters are collected every 5ms to monitor whether the anomaly has been eliminated;

[0031] S4.2 Effect Judgment Unit: If the parameters return to the normal range within 100ms, the protection is deemed effective; if it is not effective, the protection strategy is upgraded (minor → moderate → severe).

[0032] S4.3 Fault Reporting Unit: Reports data such as anomaly type, protective action, and execution effect to the vehicle controller (VCU) and stores it in the local data module for a storage period of ≥1 year;

[0033] (5) Human-computer interaction and data traceability module

[0034] S5.1 Alarm Unit: Pushes abnormality level and protective action prompts to the driver through instruments and voice;

[0035] S5.2 Data storage unit: Stores data for the entire process of anomaly identification, protection actions, and system feedback, with a data sampling frequency of 10kHz and a storage capacity of ≥16GB;

[0036] S5.3 Data Reading Unit: Supports reading stored data via OBD interface, facilitating fault analysis and tracing.

[0037] More preferably, the random forest algorithm of the anomaly recognition model unit contains 50 to 100 decision trees, and the model training dataset covers 100,000+ high-voltage electrical anomaly conditions, ensuring an accuracy rate of ≥99%.

[0038] More preferably, the sensor unit of the multi-dimensional data acquisition module adopts a redundant design, with both voltage and current sensors having primary and backup channels. When the primary sensor fails, it automatically switches to the backup sensor with a switching delay of ≤5ms.

[0039] Furthermore, the protection strategy execution unit of the graded protection control module supports custom configuration, and can adjust the graded threshold and protection action according to the high voltage architecture of different vehicle models, adapting to 400V / 800V voltage levels.

[0040] In a further preferred embodiment, the effect judgment unit of the closed-loop feedback module is equipped with a timeout mechanism. If the abnormality is not eliminated within 500ms after the protection action is executed, a severe abnormality protection action (emergency power cut-off) is directly triggered.

[0041] Furthermore, the alarm unit of the human-computer interaction and data traceability module supports multi-level alarms: minor abnormalities are indicated only by instrument display, moderate abnormalities are indicated by instrument and voice prompts, and severe abnormalities are indicated by instrument, voice, and light alarms.

[0042] More preferably, the Kalman filtering algorithm of the multi-dimensional data acquisition module can eliminate data fluctuations caused by electromagnetic interference, and the data fluctuation amplitude after filtering is ≤±0.5%.

[0043] Furthermore, the feature extraction unit of the abnormal state recognition module can dynamically adjust the feature weights according to the vehicle's operating conditions (fast charging / slow charging / driving / parking) to improve the recognition accuracy under different operating conditions.

[0044] In a further preferred embodiment, the action execution drive unit of the graded protection control module adopts hard-wire priority control logic. When the CAN bus fails, the hard wire can still trigger an emergency power-off action to ensure that the safety protection does not fail.

[0045] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions:

[0046] I. This invention integrates multi-dimensional parameters to construct an anomaly identification model, which can identify latent anomalies such as insulation degradation and increased contact resistance 5-10 seconds in advance. The fault response time is ≤100ms, avoiding the risk of fault escalation from the source. The occurrence rate of high-voltage electrical faults in the whole vehicle is reduced by more than 80%. The abnormal state is divided into three levels: minor, moderate and severe, corresponding to the graded protection actions of "parameter adjustment - power reduction operation - emergency power cut-off". This avoids minor anomalies triggering the whole vehicle power cut-off, reduces the life loss of the power battery caused by frequent power cut-offs by 30%, and significantly improves the driving experience.

[0047] Second, this invention integrates more than 6 types of monitoring parameters based on machine learning algorithms, with an anomaly identification accuracy of ≥99% and a false judgment rate of ≤1%, significantly reducing ineffective protection actions and reducing after-sales maintenance costs by 40%. After the protection action is executed, the system status is monitored in real time. If the fault is not eliminated, the protection strategy is upgraded to ensure that the protection measures are 100% effective and eliminate the risk of "protection failure".

[0048] Third, this invention adopts a modular architecture, supports vehicle adaptation for different voltage levels (400V / 800V) and different topologies, shortens the adaptation cycle to 1 month, reduces development costs by 50%, and has a built-in data storage module to record data on the entire process of anomaly identification, protection actions, and system feedback, enabling the tracing of the root cause of the fault and providing data support for the design optimization of high-voltage electrical systems.

[0049] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a flowchart of the system operation of the present invention. Detailed Implementation

[0052] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0053] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0054] like Figure 1 As shown, this embodiment of the invention provides a protection and control system based on the identification of abnormal conditions in vehicle-mounted high-voltage electrical systems, including the following modules:

[0055] (1) Multi-dimensional data acquisition module

[0056] S1.1 Sensor Unit: Includes voltage sensor (measurement range 0~1000V, accuracy ±0.5%), current sensor (measurement range 0~500A, accuracy ±0.1%), temperature sensor (measurement range -40~150℃, accuracy ±1℃), insulation resistance sensor (measurement range 0~1000MΩ, accuracy ±2%), and vibration sensor (measurement range 0~2000Hz, accuracy ±0.1g), which are deployed at key nodes such as the power battery output terminal, high voltage distribution box, motor controller, and charging interface, respectively.

[0057] S1.2 Data Preprocessing Unit: Filters (using Kalman filter algorithm), denoises, and normalizes the collected raw data, with a sampling frequency ≥10kHz to ensure data accuracy;

[0058] S1.3 Data transmission unit: The pre-processed data is transmitted to the anomaly identification module using a CANFD bus (transmission rate 8Mbps), with a transmission delay of ≤10ms;

[0059] (2) Abnormal state identification module

[0060] S2.1 Feature Extraction Unit: Extracts 15 types of feature parameters, including voltage fluctuation coefficient, current harmonic content, temperature change rate, insulation impedance attenuation rate, and vibration spectrum characteristics;

[0061] S2.2 Anomaly Identification Model Unit: An anomaly identification model is constructed based on the random forest algorithm. The model input consists of 15 types of feature parameters, and the output consists of the anomaly status level (slight / moderate / severe) and anomaly type (insulation degradation / overcurrent / overvoltage / poor contact, etc.).

[0062] S2.3 Real-time Judgment Unit: The identification result is updated every 10ms, and a warning is given 5-10 seconds in advance for latent anomalies (such as insulation resistance decreasing from 500MΩ to 200MΩ). The anomaly identification accuracy is ≥99%.

[0063] (3) Graded protection control module

[0064] S3.1 Grading Determination Unit: Preset Abnormal State Grading Threshold:

[0065] Minor abnormalities: A single parameter exceeds the normal range by less than 10%, such as voltage fluctuation ±5% or insulation resistance ≥200MΩ;

[0066] Moderate abnormality: Multiple parameters exceed the normal range by 10% to 30%, such as voltage fluctuation of ±10% and temperature increase of 5℃, and insulation resistance of 100~200MΩ;

[0067] Severe abnormality: Parameters exceed the normal range by more than 30% or dangerous operating conditions occur, such as overcurrent ≥150% of rated current, insulation resistance ≤100MΩ;

[0068] S3.2 Protection Strategy Execution Unit: Executes corresponding protection actions based on the anomaly level:

[0069] Minor anomalies: Adjust the power battery output voltage / current and cooling system power to bring the parameters back to the normal range;

[0070] Moderate abnormality: Control the motor controller to reduce power to 50% of rated power and limit fast charging current to 30% of rated current;

[0071] Severe anomaly: Triggers an emergency power outage of the high-voltage distribution box, cutting off the high-voltage circuit of the entire vehicle and activating the vehicle alarm system;

[0072] S3.3 Action Execution Drive Unit: Through hard-wired + CAN dual-channel control, it ensures a 100% success rate in executing protective actions with an execution delay of ≤50ms;

[0073] (4) Closed-loop feedback module

[0074] S4.1 Status Monitoring Unit: After the protective action is executed, key system parameters are collected every 5ms to monitor whether the anomaly has been eliminated;

[0075] S4.2 Effect Judgment Unit: If the parameters return to the normal range within 100ms, the protection is deemed effective; if it is not effective, the protection strategy is upgraded (minor → moderate → severe).

[0076] S4.3 Fault Reporting Unit: Reports data such as anomaly type, protective action, and execution effect to the vehicle controller (VCU) and stores it in the local data module for a storage period of ≥1 year;

[0077] (5) Human-computer interaction and data traceability module

[0078] S5.1 Alarm Unit: Pushes abnormality level and protective action prompts to the driver through instruments and voice;

[0079] S5.2 Data storage unit: Stores data for the entire process of anomaly identification, protection actions, and system feedback, with a data sampling frequency of 10kHz and a storage capacity of ≥16GB;

[0080] S5.3 Data Reading Unit: Supports reading stored data via OBD interface, facilitating fault analysis and tracing.

[0081] In one embodiment, the random forest algorithm of the anomaly identification model unit contains 50 to 100 decision trees, and the model training dataset covers 100,000+ high-voltage electrical anomaly conditions to ensure an identification accuracy of ≥99%.

[0082] In one embodiment, the sensor unit of the multi-dimensional data acquisition module adopts a redundant design, with both voltage and current sensors configured as primary and backup. When the primary sensor fails, it automatically switches to the backup sensor with a switching delay of ≤5ms.

[0083] In one embodiment, the protection strategy execution unit of the graded protection control module supports custom configuration, and can adjust the graded threshold and protection action according to the high voltage architecture of different vehicle models, adapting to 400V / 800V voltage levels.

[0084] In one embodiment, the effect judgment unit of the closed-loop feedback module is set with a timeout mechanism. If the abnormality is not eliminated within 500ms after the protection action is executed, the severe abnormality protection action (emergency power failure) is directly triggered.

[0085] In one embodiment, the alarm unit of the human-computer interaction and data traceability module supports multi-level alarms: minor abnormalities are indicated only by instrument display, moderate abnormalities are indicated by instrument and voice prompts, and severe abnormalities are indicated by instrument, voice, and light alarms.

[0086] In one embodiment, the Kalman filter algorithm of the multi-dimensional data acquisition module can eliminate data fluctuations caused by electromagnetic interference, and the data fluctuation amplitude after filtering is ≤±0.5%.

[0087] In one embodiment, the feature extraction unit of the abnormal state recognition module can also dynamically adjust the feature weights according to the vehicle's operating conditions (fast charging / slow charging / driving / parking) to improve the recognition accuracy under different operating conditions.

[0088] In one embodiment, the action execution drive unit of the graded protection control module adopts hard-wire priority control logic. When the CAN bus fails, the hard wire can still trigger an emergency power-off action to ensure that the safety protection does not fail.

[0089] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A protection and control system based on vehicle-mounted high-voltage electrical anomaly identification, characterized in that: Includes the following modules: (1) Multi-dimensional data acquisition module S1.1 Sensor Unit: Includes voltage sensor (measurement range 0~1000V, accuracy ±0.5%), current sensor (measurement range 0~500A, accuracy ±0.1%), temperature sensor (measurement range -40~150℃, accuracy ±1℃), insulation resistance sensor (measurement range 0~1000MΩ, accuracy ±2%), and vibration sensor (measurement range 0~2000Hz, accuracy ±0.1g), which are deployed at key nodes such as the power battery output terminal, high voltage distribution box, motor controller, and charging interface, respectively. S1.2 Data Preprocessing Unit: Filters (using Kalman filter algorithm), denoises, and normalizes the collected raw data, with a sampling frequency ≥10kHz to ensure data accuracy; S1.3 Data transmission unit: The pre-processed data is transmitted to the anomaly identification module using a CANFD bus (transmission rate 8Mbps), with a transmission delay of ≤10ms; (2) Abnormal state identification module S2.1 Feature Extraction Unit: Extracts 15 types of feature parameters, including voltage fluctuation coefficient, current harmonic content, temperature change rate, insulation impedance attenuation rate, and vibration spectrum characteristics; S2.2 Anomaly Identification Model Unit: An anomaly identification model is constructed based on the random forest algorithm. The model input consists of 15 types of feature parameters, and the output consists of the anomaly status level (slight / moderate / severe) and anomaly type (insulation degradation / overcurrent / overvoltage / poor contact, etc.). S2.3 Real-time Judgment Unit: The identification result is updated every 10ms, and a warning is given 5-10 seconds in advance for latent anomalies (such as insulation resistance decreasing from 500MΩ to 200MΩ). The anomaly identification accuracy is ≥99%. (3) Graded protection control module S3.1 Grading Determination Unit: Preset Abnormal State Grading Threshold: Minor abnormalities: A single parameter exceeds the normal range by less than 10%, such as voltage fluctuation ±5% or insulation resistance ≥200MΩ; Moderate abnormality: Multiple parameters exceed the normal range by 10% to 30%, such as voltage fluctuation of ±10% and temperature increase of 5℃, and insulation resistance of 100~200MΩ; Severe abnormality: Parameters exceed the normal range by more than 30% or dangerous operating conditions occur, such as overcurrent ≥150% of rated current, insulation resistance ≤100MΩ; S3.2 Protection Strategy Execution Unit: Executes corresponding protection actions based on the anomaly level: Minor anomalies: Adjust the power battery output voltage / current and cooling system power to bring the parameters back to the normal range; Moderate abnormality: Control the motor controller to reduce power to 50% of rated power and limit fast charging current to 30% of rated current; Severe anomaly: Triggers an emergency power outage of the high-voltage distribution box, cutting off the high-voltage circuit of the entire vehicle and activating the vehicle alarm system; S3.3 Action Execution Drive Unit: Through hard-wired + CAN dual-channel control, it ensures a 100% success rate in executing protective actions with an execution delay of ≤50ms; (4) Closed-loop feedback module S4.1 Status Monitoring Unit: After the protective action is executed, key system parameters are collected every 5ms to monitor whether the anomaly has been eliminated; S4.2 Effect Judgment Unit: If the parameters return to the normal range within 100ms, the protection is deemed effective; if it is not effective, the protection strategy is upgraded (minor → moderate → severe). S4.3 Fault Reporting Unit: Reports data such as anomaly type, protective action, and execution effect to the vehicle controller (VCU) and stores it in the local data module for a storage period of ≥1 year; (5) Human-computer interaction and data traceability module S5.1 Alarm Unit: Pushes abnormality level and protective action prompts to the driver through instruments and voice; S5.2 Data storage unit: Stores data for the entire process of anomaly identification, protection actions, and system feedback, with a data sampling frequency of 10kHz and a storage capacity of ≥16GB; S5.3 Data Reading Unit: Supports reading stored data via OBD interface, facilitating fault analysis and tracing.

2. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The anomaly identification model unit uses a random forest algorithm containing 50 to 100 decision trees, and the model training dataset covers 100,000+ high-voltage electrical anomaly conditions, ensuring an identification accuracy of ≥99%.

3. The protection and control system based on vehicle-mounted high-voltage electrical anomaly identification according to claim 1, characterized in that: The sensor unit of the multi-dimensional data acquisition module adopts a redundant design. The voltage / current sensors are equipped with two channels, one for primary and one for backup. When the primary sensor fails, it automatically switches to the backup sensor with a switching delay of ≤5ms.

4. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The protection strategy execution unit of the graded protection control module supports custom configuration and can adjust the graded threshold and protection action according to the high voltage architecture of different vehicle models, adapting to 400V / 800V voltage levels.

5. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The effect judgment unit of the closed-loop feedback module is set with a timeout mechanism. If the abnormality is not eliminated within 500ms after the protection action is executed, the severe abnormality protection action (emergency power cut) is directly triggered.

6. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The alarm unit of the human-computer interaction and data traceability module supports multi-level alarms: minor abnormalities are indicated only by instrument display, moderate abnormalities are indicated by instrument and voice prompts, and severe abnormalities are indicated by instrument, voice, and light alarms.

7. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The Kalman filter algorithm of the multi-dimensional data acquisition module can eliminate data fluctuations caused by electromagnetic interference, and the data fluctuation amplitude after filtering is ≤±0.5%.

8. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The feature extraction unit of the abnormal state recognition module can also dynamically adjust the feature weights according to the vehicle's operating conditions (fast charging / slow charging / driving / parking) to improve the recognition accuracy under different operating conditions.

9. The protection and control system based on vehicle-mounted high-voltage electrical abnormality identification according to claim 1, characterized in that: The action execution drive unit of the graded protection control module adopts hard-wire priority control logic. When the CAN bus fails, the hard wire can still trigger an emergency power-off action to ensure that the safety protection does not fail.

Images