An automatic driving-oriented low-voltage power supply method and system, and a vehicle

Abstract

The application provides an automatic driving-oriented low-voltage power supply method and system and a vehicle, and belongs to the technical field of electric vehicle power supply. The application comprises the following steps: obtaining working state information of a high-voltage power system; determining a current power supply mode according to the working state information of the high-voltage power system; when the high-voltage power system is in a normal working state, controlling a direct-current converter to serve as a main power supply to supply power to a whole-vehicle low-voltage load and simultaneously charge a main lithium battery and a redundant lithium battery; when the high-voltage power system is in a fault state, controlling the main lithium battery and the redundant lithium battery to enter an independent power supply mode and respectively provide double backup power for the whole-vehicle low-voltage load; and continuously detecting the states of each power supply loop through an isolation protection module, cutting off an abnormal loop when detecting that any power supply loop is abnormal, and keeping the remaining loops continuously supplying power. The application avoids systematic risks caused by single-point power supply failure.

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle power supply technology, and particularly relates to a low-voltage power supply method, system and vehicle for autonomous driving. Background Technology

[0002] Currently, electric vehicle low-voltage power systems generally use a single 12V lithium battery as the low-voltage energy storage unit, which is charged by a DC-DC converter on the high-voltage power battery side and simultaneously supplies power to the vehicle's low-voltage equipment. When the system exits Level 3 autonomous driving mode and requests driver takeover, the low-voltage power supply needs to continuously supply power to critical equipment for a certain period of time to support the user in completing driving operations.

[0003] However, the aforementioned single-battery architecture carries a single point of failure risk: when an anomaly occurs in the low-voltage power supply circuit, the entire vehicle's low-voltage system faces the risk of power outage. A single lithium-ion battery solution struggles to maintain the continuous operation of critical low-voltage loads when the power supply link fails, failing to meet the reliability and continuous power supply requirements of Level 3 autonomous driving for low-voltage power systems. Furthermore, existing solutions can only respond passively after a fault occurs, unable to implement preventative switching before the fault is formally triggered, posing a risk of momentary power outage. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, this invention proposes a low-voltage power supply method, system, and vehicle for autonomous driving, in order to solve the technical problems of single-point failure risk, lack of redundant power supply capability, and inability to predictively switch fault trends in the existing single 12V lithium battery architecture.

[0005] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: In a first aspect, the present invention discloses a low-voltage power supply method for autonomous driving, comprising: Obtain the operating status information of the high-voltage power system; Based on the operating status information of the high-voltage power system, the current power supply mode is determined. When the high-voltage power system is in normal operating condition, the DC-DC converter is controlled as the main power supply to provide power to the low-voltage load of the vehicle, and the main lithium battery and redundant lithium battery are charged at the same time. When the high-voltage power system is in fault condition, the main lithium battery and redundant lithium battery are controlled to enter independent power supply mode, providing dual backup power to the low-voltage load of the vehicle respectively. The isolation protection module continuously monitors the status of each power supply circuit. When any abnormality is detected in any power supply circuit, the abnormal circuit is cut off, while the remaining circuits continue to be powered.

[0006] A further technical solution, the method also includes: when the autonomous driving system fails and the high-voltage power system is in normal working condition, maintaining the DC-DC converter power supply mode, issuing a takeover request to the dynamic driving task support user, and continuously supplying power to the braking system, steering system and intelligent driving system before the takeover is completed.

[0007] A further technical solution involves detecting the status of each power supply circuit. When an abnormality is detected in any power supply circuit, the abnormal circuit is disconnected, while the remaining circuits continue to receive power. Specifically, this includes: The current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence; The system detects whether there are any abnormalities such as short circuit, overvoltage, undervoltage, or open circuit based on the time-series sampling sequence; if the above abnormalities are detected, the isolation protection module is triggered to cut off the faulty branch. If no abnormality is detected, trend prediction is performed; an electrical parameter trend model is established for the time-series sampling sequence, and the electrical parameter change trend of each power supply circuit is predicted based on the trend model within a preset prediction time window. When the prediction result indicates that the electrical parameter of any power supply circuit will reach the abnormal judgment condition within the preset prediction time window, the isolation protection module is triggered in advance to perform a pre-switching action on the circuit before the actual occurrence of the fault, so as to smoothly transfer the power supply responsibility of the circuit to the healthy circuit.

[0008] A further technical solution, wherein establishing an electrical parameter trend model for the time-series sampling sequence, specifically includes: The isolation protection module continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period to form a time-series sampling sequence for each power supply circuit; the time-series sampling sequence is used to establish an electrical parameter trend model for each power supply circuit, including a linear regression model and a multinomial fitting model. The system adaptively selects between linear regression and polynomial fitting based on the root mean square value of the fitting residuals of the time-series sampling sequence. When the root mean square value of linear regression is lower than the preset residual threshold, the linear regression model is used; otherwise, the system switches to the polynomial fitting model.

[0009] A further technical solution, the trend model prediction includes: Linear regression model extrapolation: Substitute the slope of the trend model of linear regression and the measured values ​​of electrical parameters at the current moment into the linear equation to calculate the predicted value at the end of the preset prediction time window; if the predicted value exceeds the fault judgment range, it is determined that the power supply circuit has a fault trend. Polynomial fitting model extrapolation: When a polynomial fitting model is used, each time between the current time and the end time of the preset prediction time window is substituted into the fitted polynomial equation in turn to calculate the predicted value sequence of electrical parameters within the prediction time window; the maximum and minimum values ​​in the predicted value sequence are taken as the judgment criteria. If the extreme value touches the fault judgment boundary, it is determined that the power supply circuit has a fault trend.

[0010] A further technical solution is that the low-voltage load of the vehicle includes high-safety-level loads and non-critical loads. The high-safety-level loads are electrically connected to both the main lithium battery and the redundant lithium battery, and are powered by dual-circuit redundant power supply. The non-critical loads are powered by a single circuit. The high-safety-level load includes at least one of an intelligent driving system, an electric power steering system, and an electronic braking system.

[0011] Secondly, this invention discloses a low-voltage power supply system for autonomous driving, including a high-voltage power battery and a DC-DC converter; a main lithium battery electrically connected to the DC-DC converter for receiving the charging current output by the DC-DC converter and discharging power to the low-voltage load of the vehicle; a redundant lithium battery electrically connected to the DC-DC converter for receiving the charging current output by the DC-DC converter and discharging power to the low-voltage load of the vehicle when the main lithium battery or its power supply link fails; and an isolation protection module integrated inside the DC-DC converter and signal-connected to the DC-DC converter for real-time detection of the status of each power supply circuit, cutting off the faulty circuit when any power supply circuit is detected as abnormal, and keeping the remaining circuits supplying power to the low-voltage load.

[0012] Thirdly, the present invention discloses a vehicle including the aforementioned low-voltage power supply system for autonomous driving.

[0013] Fourthly, the present invention discloses a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the aforementioned low-voltage power supply method for autonomous driving.

[0014] Fifthly, the present invention discloses a computer program product comprising executable instructions stored in a computer-readable storage medium; wherein, when the vehicle's processor reads and executes the executable instructions from the computer-readable storage medium, it completes the steps of the aforementioned low-voltage power supply method for autonomous driving.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention introduces a dual-redundant low-voltage power supply architecture consisting of a main lithium battery and redundant lithium batteries, completely eliminating the single-point-of-failure risk inherent in the existing single 12V lithium battery architecture. When any power supply path experiences a short circuit, overvoltage, undervoltage, or open circuit, the system can automatically disconnect the faulty circuit through the isolation protection module and maintain continuous power supply to the low-voltage load through the remaining circuits, overcoming the defect of the existing architecture where the entire vehicle's low-voltage system loses power when a power supply link fails.

[0016] Furthermore, by introducing a predictive fault switching mechanism, the current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence. An electrical parameter trend model is established on the time-series sampling sequence. Based on the trend model, the electrical parameter change trend of each power supply circuit is predicted by extrapolation. When the fault judgment condition is predicted to be triggered, the pre-switching action is performed in advance, advancing the system response time window to before the fault is officially triggered, effectively avoiding the risk of power outage. Compared with the traditional passive response switching, it has a significant safety improvement.

[0017] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0018] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0019] Figure 1 This is a flowchart of the low-voltage power supply method for autonomous driving as described in Embodiment 1 of the present invention.

[0020] Figure 2 This is a flowchart illustrating the power supply process of DC-DC converter and dual lithium batteries after the autonomous driving system fails, as described in Embodiment 1 of the present invention.

[0021] Figure 3 This is a schematic diagram of a low-voltage power supply system for autonomous driving as described in Embodiment 2 of the present invention. Detailed Implementation

[0022] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0023] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0024] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0025] Example 1 In one or more embodiments, a low-voltage power supply method for autonomous driving is disclosed, such as... Figure 1 As shown, it includes the following steps: Step S1: Obtain the operating status information of the autonomous driving system and the working status information of the high-voltage power system. The on-board controller collects the above status information in real time through the communication interface with the autonomous driving domain controller and the high-voltage management system, providing a data foundation for subsequent power supply mode decisions.

[0026] Step S2: Determine the current power supply mode based on the operating status information of the high-voltage power system. When the high-voltage power system is in normal operating condition, control the DC-DC converter as the main power supply to power the low-voltage load of the vehicle, and simultaneously charge the main lithium battery and the redundant lithium battery. At this time, the DC-DC converter continuously replenishes the two sets of lithium batteries through the dual-channel isolated output terminal to keep them in a fully charged standby state. When the high-voltage power system is in a fault state, control the main lithium battery and the redundant lithium battery to enter an independent power supply mode, providing dual-channel backup power to the low-voltage load of the vehicle, thereby responding to extreme conditions such as high-voltage power failure triggered by collision, serious power battery failure, or main relay disconnection. The independent power supply of the dual lithium batteries ensures that the vehicle remains in a basically controllable state.

[0027] Preferably, determining the current power supply mode further includes: such as Figure 2 As shown, when the autonomous driving system fails but the high-voltage power system is operating normally, the DC-DC converter maintains its power supply mode and sends a takeover request to the dynamic driving task support user. Power is continuously supplied to the braking system, steering system, and intelligent driving system until the takeover is completed. This step ensures that in scenarios where the autonomous driving system fails but the high-voltage power system is functioning normally, the system can continuously supply power to safety-critical loads via the DC-DC converter, meeting the mandatory requirements of GB / T 40429-2021 for continuous power supply after L3 level autonomous driving failure.

[0028] Step S3: The isolation protection module continuously monitors the status of each power supply circuit. When any abnormality is detected in any power supply circuit, the abnormal circuit is cut off, while the remaining circuits continue to be powered.

[0029] Specifically, detecting whether an abnormality has occurred in the power supply circuit includes: First, the current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence.

[0030] Secondly, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits based on the time-series sampling sequence. If such anomalies are detected, the isolation protection module is triggered within milliseconds to cut off the faulty branch while maintaining the healthy circuit to continue supplying power to the high-safety-level load, preventing the fault from spreading and ensuring the system's functional safety. If no such anomalies are detected, trend prediction is performed. Then, an electrical parameter trend model is established for the time-series sampling sequence. Based on the trend model, the electrical parameter change trend of each power supply circuit within a preset prediction time window is predicted. When the prediction result shows that the electrical parameter of any power supply circuit will reach the abnormal judgment condition within the preset prediction time window, the isolation protection module is triggered in advance to perform a pre-switching action on the circuit before the actual occurrence of the fault, so as to smoothly transfer the power supply responsibility of the circuit to the healthy circuit.

[0031] Furthermore, based on the time-series sampling sequence, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits. Specifically, it compares the current and voltage values ​​of adjacent sampling points in the time-series sampling sequence in real time and identifies four types of anomalies based on the following characteristics: (1) Short circuit anomaly: The current value suddenly increases between adjacent sampling points and the voltage value drops in the same step, indicating that a low impedance short circuit fault has occurred in the power supply circuit; preferably, within a single sampling period, the increase in the current value exceeds a preset multiple of the rated operating current, which is set to 3 times in this embodiment, and is thus judged as a sudden increase in current; at the same time, within the same sampling period, the drop in the voltage value exceeds a preset proportion of the rated operating voltage, which is taken as 20%, and the short circuit anomaly is judged when both conditions are met.

[0032] (2) Overvoltage abnormality: The voltage value continuously exceeds the upper limit of the normal working range, indicating that the power supply circuit is overcharged or the voltage regulation fails; preferably, in this embodiment, the voltage values ​​of three or more consecutive sampling points continuously exceed the upper limit of the normal working range.

[0033] (3) Undervoltage abnormality: The voltage value is consistently lower than the lower limit of the normal operating range, indicating that the battery capacity is insufficient or there is a high impedance fault in the power supply link.

[0034] (4) Open circuit anomaly: The current value drops sharply to near zero and the voltage value rises abnormally, indicating that an open circuit fault has occurred in the power supply circuit. Specifically, if the current value drops by more than a preset proportion of the rated operating current within a single sampling period (set to 90%) and the current value after the drop is lower than the preset lower limit of the rated operating current (set to 5%), it is judged as a sharp drop in current to near zero; if the voltage value increases by more than a preset proportion of the rated operating voltage within the same sampling period (set to 15%), it is judged as an abnormal voltage rise; an open circuit anomaly is judged only when both conditions are met simultaneously.

[0035] When any of the above features are detected, it is determined that a corresponding abnormality has occurred in the power supply circuit. Within milliseconds, the isolation protection module is triggered to disconnect the faulty branch, while the healthy circuit continues to supply power to the high-safety-level load, preventing the fault from spreading and ensuring the functional safety of the system.

[0036] It should be understood that the normal operating range is calibrated during system initialization based on the battery rating parameters and the vehicle's low-voltage system specifications, or pre-configured at the factory.

[0037] Furthermore, predictive failover includes the following steps: (1) Trend modeling: The isolation protection module continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period to form a time-series sampling sequence of each power supply circuit; the time-series sampling sequence is used to establish an electrical parameter trend model of each power supply circuit by using linear regression or polynomial fitting method, so as to provide a data basis for fault prediction.

[0038] Specifically, linear regression modeling: using the timestamp of each sampling point in the time-series sampling sequence as the independent variable and the corresponding current value or voltage value as the dependent variable, the least squares method is used to fit a linear equation to obtain the slope of the electrical parameters changing with time; the sign of the slope reflects the direction of parameter change, and the absolute value of the slope reflects the rate of parameter degradation. When the absolute value of the slope continues to increase, it indicates that there is an accelerated degradation trend in the power supply circuit.

[0039] Polynomial fitting modeling: When electrical parameters exhibit a nonlinear degradation trend, a quadratic or cubic polynomial is used to fit the time-series sampling sequence to more accurately capture the curvature characteristics of accelerated degradation. The quadratic polynomial is suitable for uniformly accelerated degradation scenarios, while the cubic polynomial is suitable for complex scenarios where the degradation rate itself is also changing.

[0040] Furthermore, the model selection strategy is as follows: linear regression or polynomial fitting is adaptively selected based on the root mean square value (RMSE) of the fitting residuals of the time-series sampling sequence. When the RMSE of linear regression is lower than the preset residual threshold, the linear regression model is used; otherwise, the polynomial fitting model is switched to achieve a balance between computational cost and fitting accuracy.

[0041] In this embodiment, the preset residual threshold can be calibrated according to the actual parameters of the vehicle's low-voltage system. For example, for current signals, the preset residual threshold can be set to 0.5A; for voltage signals, the preset residual threshold can be set to 0.1V. When the RMSE of the linear regression model is lower than the above-mentioned preset residual threshold, the linear regression model is used; otherwise, the model is switched to a multinomial fitting model.

[0042] The above modeling process is updated on a rolling basis in each sampling period. The trend model is updated once a new sampling point is acquired, ensuring that the trend model always reflects the latest state of the power supply circuit and provides a real-time data basis for fault prediction.

[0043] (2) Trend prediction and early warning: Based on the trend model, the electrical parameter change trend of each power supply circuit is predicted within the preset prediction time window; when the prediction result shows that the electrical parameter of any power supply circuit will reach the fault judgment condition within the preset prediction time window, the system determines that the power supply circuit has a fault trend and issues an early warning signal.

[0044] The specific process is as follows: Linear regression model extrapolation: Substitute the slope of the trend model of linear regression and the measured values ​​of electrical parameters at the current moment into the linear equation to calculate the predicted value at the end of the preset prediction time window; if the predicted value exceeds the fault judgment range, it is determined that there is a fault trend in the power supply circuit and an early warning signal is issued. Polynomial fitting model extrapolation: When a polynomial fitting model is used, each time between the current time and the end time of the preset prediction time window is substituted into the fitted polynomial equation in turn to calculate the predicted value sequence of electrical parameters within the prediction time window; the maximum and minimum values ​​in the predicted value sequence are taken as the judgment criteria. If the extreme value touches the fault judgment boundary, it is determined that the power supply circuit has a fault trend and a warning signal is issued. The upper limit of the rated operating current and the upper and lower limits of the rated operating voltage of the power supply circuit are used as the fault judgment boundary; when the extrapolated predicted value touches the above boundary within the preset prediction time window, it is determined that the power supply circuit has a fault trend. Preferably, the prediction time window can be differentiated. For power supply circuits corresponding to high safety level loads, a relatively longer preset prediction time window (e.g., 100ms) can be used to trigger the pre-switching action with a greater time lead, ensuring uninterrupted power supply to high safety level loads. For power supply circuits corresponding to non-critical loads, a relatively shorter prediction time window (e.g., 20ms) can be used to avoid frequent false triggering of pre-switching actions due to prediction errors.

[0045] Compared with existing passive response switching, predictive early warning based on trend models can advance the system response time window to before the fault is officially triggered, effectively avoiding the risk of power outage caused by fault current surges.

[0046] Preferably, the power supply ratio of the healthy circuit is gradually increased within a preset switching time window (e.g., 5ms) in a gradual manner, while the power supply ratio of the circuit with a fault trend is gradually decreased. The sum of the power supply ratios of the two circuits remains at 100% throughout the entire switching process, ensuring that the load end receives a continuous and uninterrupted power supply.

[0047] This predictive switching mechanism ensures the continuity of power supply to the vehicle's low-voltage loads before a fault actually occurs, which is of great significance for ensuring uninterrupted power supply to high-safety-level loads such as intelligent driving systems, electric power steering systems, and electronic braking systems.

[0048] Example 2 In one or more embodiments, a low-voltage power supply system for autonomous driving is disclosed, such as... Figure 3 As shown, it specifically includes a high-voltage power battery, a DC-DC converter, a main lithium battery, redundant lithium batteries, an isolation protection module, and a fuse box.

[0049] The high-voltage power battery, as the high-voltage energy storage unit of the vehicle, is electrically connected to the high-voltage input terminal of the DC-DC converter through a high-voltage wiring harness, providing high-voltage DC power input to the DC-DC converter.

[0050] The DC-DC converter converts the high-voltage DC output from the high-voltage power battery into low-voltage DC, and outputs charging current to the main lithium battery and redundant lithium battery respectively through dual isolated output terminals, while simultaneously supplying power to the low-voltage load of the vehicle. The DC-DC converter has dual isolated output terminals, with the first output terminal electrically connected to the main lithium battery and the second output terminal electrically connected to the redundant lithium battery. The two output terminals are electrically isolated from each other, ensuring that a failure in one terminal does not affect the normal operation of the other terminal.

[0051] The main lithium battery has a nominal voltage of 48V and is electrically connected to the first output terminal of the DC-DC converter. It receives the charging current output by the DC-DC converter and is in a charging maintenance state when the high-voltage power system is working normally, while simultaneously discharging power to the low-voltage load of the vehicle. Compared with traditional 12V lithium batteries, the 48V main lithium battery has higher energy density and larger energy storage capacity, providing a longer backup power time for the vehicle after the high-voltage power system fails.

[0052] The redundant lithium battery has a nominal voltage of 12V and is electrically connected to the second output of the DC-DC converter. It receives the charging current from the DC-DC converter and automatically assumes the responsibility of discharging power to the vehicle's low-voltage loads in the event of a failure in the main lithium battery or its power supply link, forming a dual-redundancy protection system. The 12V redundant lithium battery is compatible with the electrical interfaces of existing vehicle low-voltage equipment, facilitating system integration.

[0053] This invention employs a dual-battery configuration of a 48V main lithium battery and a 12V redundant lithium battery, which significantly improves the backup power supply time and total energy storage capacity after the failure of the high-voltage power system compared to the traditional single 20Ah lithium battery solution. This meets the time requirement of L3 autonomous driving to continuously supply power to critical low-voltage loads until the driver takes over after the autonomous driving function is discontinued.

[0054] The isolation protection module is integrated inside the DC-DC converter, connected to the DC-DC converter signal, and designed in accordance with ASIL B functional safety level. The isolation protection module has two operating modes: The first is a passive fault isolation mode, used to monitor the voltage, current, and insulation status of each power supply circuit in real time. When any power supply circuit experiences a short circuit, overvoltage, undervoltage, or open circuit anomaly, the faulty circuit is disconnected within milliseconds to achieve electrical isolation, while maintaining continuous power supply to the low-voltage load from the remaining healthy circuits. The second is a predictive fault switching mode, which continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period (e.g., 1ms) to form a time-series sampling sequence. A linear regression or polynomial fitting method is used to establish a trend model of the electrical parameters of each power supply circuit. Based on this trend model, the module extrapolates forward to predict the change trend of electrical parameters of each power supply circuit within a preset prediction time window. When the prediction result indicates that the electrical parameters of any power supply circuit will reach the fault judgment condition within the preset prediction time window, a pre-switching action is performed before the actual fault occurs, smoothly transferring the power supply responsibility to the healthy circuit. For power supply circuits corresponding to high-safety-level loads, a relatively longer preset prediction time window is used to trigger the pre-switching action with a greater time lead, ensuring uninterrupted power supply to high-safety-level loads.

[0055] This invention integrates the isolation protection module into the DC-DC converter and is designed in accordance with ASIL B functional safety level, realizing continuous monitoring of voltage, current and insulation status of each power supply circuit and millisecond-level fault isolation, effectively preventing the spread of local faults, and prioritizing the continuous power supply of high safety level loads in extreme scenarios such as autonomous driving failure, which meets the mandatory requirements of GB / T 40429-2021 for the functional safety of L3 level autonomous driving low-voltage power supply systems.

[0056] The fuse box is electrically connected to the main lithium battery and the redundant lithium battery respectively, and is used to distribute the power output from the main lithium battery and the redundant lithium battery to the low-voltage electrical equipment of the vehicle, so as to realize centralized power distribution and overcurrent protection.

[0057] In some implementations, the low-voltage loads of the vehicle include high-safety-level loads and non-critical loads. High-safety-level loads are electrically connected to both the main lithium battery and redundant lithium batteries, employing dual-circuit redundant power supply; non-critical loads are powered by a single circuit. Specifically, high-safety-level loads include at least one of the intelligent driving system, electric power steering system, and electronic braking system. These loads directly relate to the vehicle's ability to safely take over after autonomous driving failure; implementing dual-circuit redundant power supply ensures the vehicle remains basically controllable until the driver takes over. Using a single circuit for non-critical loads effectively reduces the overall system complexity and cost while ensuring the reliability of power supply to high-safety-level loads.

[0058] In some implementations, both the main lithium battery and the redundant lithium battery are installed in the collision safety zone of the passenger compartment, and are located close to the fuse box. Specifically, the collision safety zone is the space under the seats, including under the driver and front passenger seats or under the second-row seats. This space is protected by the vehicle body structure and offers good safety protection in a collision, effectively avoiding mechanical damage to the battery caused by a vehicle collision. Located close to the fuse box, it shortens the power supply harness length, reducing line impedance and the risk of harness failure.

[0059] In some implementations, when a high-voltage power battery or DC-DC converter fails, causing the DC-DC converter to stop working, the main lithium battery and redundant lithium batteries enter an independent power supply mode, providing dual backup power to the vehicle's low-voltage loads. In this mode, even if a single lithium battery fails, the system can still rely on the other battery to maintain power supply to critical loads, significantly improving the system's continuous operating capability under extreme conditions.

[0060] The above system executes the following power supply control process: Step 1: Acquire the operational status information of the autonomous driving system and the working status information of the high-voltage power system. The on-board controller collects the above status information in real time through the communication interface with the autonomous driving domain controller and the high-voltage management system, providing a data foundation for subsequent power supply mode decisions.

[0061] Step 2: Determine the current power supply mode based on the operating status information of the high-voltage power system. When the high-voltage power system is in normal operating condition, control the DC-DC converter as the main power supply to power the low-voltage load of the vehicle, and simultaneously charge the main lithium battery and the redundant lithium battery. At this time, the DC-DC converter continuously replenishes the two sets of lithium batteries through the dual-channel isolated output terminal to keep them in a fully charged standby state. When the high-voltage power system is in a fault state, control the main lithium battery and the redundant lithium battery to enter an independent power supply mode, providing dual-channel backup power to the low-voltage load of the vehicle, thereby responding to extreme conditions such as high-voltage power failure triggered by collision, serious power battery failure, or main relay disconnection. The independent power supply of the dual lithium batteries ensures that the vehicle remains in a basically controllable state.

[0062] Preferably, determining the current power supply mode further includes: when the autonomous driving system fails but the high-voltage power system is operating normally, maintaining the DC-DC converter power supply mode, and issuing a takeover request to the dynamic driving task support user, continuously supplying power to the braking system, steering system, and intelligent driving system until the takeover is completed. This step ensures that in scenarios where the autonomous driving system fails but the high-voltage power system is normal, the system can rely on the DC-DC converter to continuously supply power to safety-critical loads, meeting the mandatory requirements of GB / T 40429-2021 for continuous power supply after L3 level autonomous driving failure.

[0063] Step 3: The isolation protection module continuously monitors the status of each power supply circuit. When any abnormality is detected in any power supply circuit, the abnormal circuit is cut off, while the remaining circuits continue to be powered.

[0064] Specifically, detecting whether an abnormality has occurred in the power supply circuit includes: First, the current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence.

[0065] Secondly, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits based on the time-series sampling sequence. If such anomalies are detected, the isolation protection module is triggered within milliseconds to cut off the faulty branch while maintaining the healthy circuit to continue supplying power to the high-safety-level load, preventing the fault from spreading and ensuring the system's functional safety. If no such anomalies are detected, trend prediction is performed. Then, an electrical parameter trend model is established for the time-series sampling sequence. Based on the trend model, the electrical parameter change trend of each power supply circuit within a preset prediction time window is predicted. When the prediction result shows that the electrical parameter of any power supply circuit will reach the abnormal judgment condition within the preset prediction time window, the isolation protection module is triggered in advance to perform a pre-switching action on the circuit before the actual occurrence of the fault, so as to smoothly transfer the power supply responsibility of the circuit to the healthy circuit.

[0066] Furthermore, based on the time-series sampling sequence, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits. Specifically, it compares the current and voltage values ​​of adjacent sampling points in the time-series sampling sequence in real time and identifies four types of anomalies based on the following characteristics: (1) Short circuit anomaly: The current value suddenly increases between adjacent sampling points and the voltage value drops in the same step, indicating that a low impedance short circuit fault has occurred in the power supply circuit; preferably, within a single sampling period, the increase in the current value exceeds a preset multiple of the rated operating current, which is set to 3 times in this embodiment, and is thus judged as a sudden increase in current; at the same time, within the same sampling period, the drop in the voltage value exceeds a preset proportion of the rated operating voltage, which is taken as 20%, and the short circuit anomaly is judged when both conditions are met.

[0067] (2) Overvoltage abnormality: The voltage value continuously exceeds the upper limit of the normal working range, indicating that the power supply circuit is overcharged or the voltage regulation fails; preferably, in this embodiment, the voltage values ​​of three or more consecutive sampling points continuously exceed the upper limit of the normal working range.

[0068] (3) Undervoltage abnormality: The voltage value is consistently lower than the lower limit of the normal operating range, indicating that the battery capacity is insufficient or there is a high impedance fault in the power supply link.

[0069] (4) Open circuit anomaly: The current value drops sharply to near zero and the voltage value rises abnormally, indicating that an open circuit fault has occurred in the power supply circuit. Specifically, if the current value drops by more than a preset proportion of the rated operating current within a single sampling period (set to 90%) and the current value after the drop is lower than the preset lower limit of the rated operating current (set to 5%), it is judged as a sharp drop in current to near zero; if the voltage value increases by more than a preset proportion of the rated operating voltage within the same sampling period (set to 15%), it is judged as an abnormal voltage rise; an open circuit anomaly is judged only when both conditions are met simultaneously.

[0070] When any of the above features are detected, it is determined that a corresponding abnormality has occurred in the power supply circuit. Within milliseconds, the isolation protection module is triggered to disconnect the faulty branch, while the healthy circuit continues to supply power to the high-safety-level load, preventing the fault from spreading and ensuring the functional safety of the system.

[0071] It should be understood that the normal operating range is calibrated during system initialization based on the battery rating parameters and the vehicle's low-voltage system specifications, or pre-configured at the factory.

[0072] Furthermore, predictive failover includes the following steps: (1) Trend modeling: The isolation protection module continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period to form a time-series sampling sequence of each power supply circuit; the time-series sampling sequence is used to establish an electrical parameter trend model of each power supply circuit by using linear regression or polynomial fitting method, so as to provide a data basis for fault prediction.

[0073] Specifically, linear regression modeling: using the timestamp of each sampling point in the time-series sampling sequence as the independent variable and the corresponding current value or voltage value as the dependent variable, the least squares method is used to fit a linear equation to obtain the slope of the electrical parameters changing with time; the sign of the slope reflects the direction of parameter change, and the absolute value of the slope reflects the rate of parameter degradation. When the absolute value of the slope continues to increase, it indicates that there is an accelerated degradation trend in the power supply circuit.

[0074] Polynomial fitting modeling: When electrical parameters exhibit a nonlinear degradation trend, a quadratic or cubic polynomial is used to fit the time-series sampling sequence to more accurately capture the curvature characteristics of accelerated degradation. The quadratic polynomial is suitable for uniformly accelerated degradation scenarios, while the cubic polynomial is suitable for complex scenarios where the degradation rate itself is also changing.

[0075] Furthermore, the model selection strategy is as follows: linear regression or polynomial fitting is adaptively selected based on the root mean square value (RMSE) of the fitting residuals of the time-series sampling sequence. When the RMSE of linear regression is lower than the preset residual threshold, the linear regression model is used; otherwise, the polynomial fitting model is switched to achieve a balance between computational cost and fitting accuracy.

[0076] In this embodiment, the preset residual threshold can be calibrated according to the actual parameters of the vehicle's low-voltage system. For example, for current signals, the preset residual threshold can be set to 0.5A; for voltage signals, the preset residual threshold can be set to 0.1V. When the RMSE of the linear regression model is lower than the above-mentioned preset residual threshold, the linear regression model is used; otherwise, the model is switched to a multinomial fitting model.

[0077] The above modeling process is updated on a rolling basis in each sampling period. The trend model is updated once a new sampling point is acquired, ensuring that the trend model always reflects the latest state of the power supply circuit and provides a real-time data basis for fault prediction.

[0078] (2) Trend prediction and early warning: Based on the trend model, the electrical parameter change trend of each power supply circuit is predicted within the preset prediction time window; when the prediction result shows that the electrical parameter of any power supply circuit will reach the fault judgment condition within the preset prediction time window, the system determines that the power supply circuit has a fault trend and issues an early warning signal.

[0079] The specific process is as follows: Linear regression model extrapolation: Substitute the slope of the trend model of linear regression and the measured values ​​of electrical parameters at the current moment into the linear equation to calculate the predicted value at the end of the preset prediction time window; if the predicted value exceeds the fault judgment range, it is determined that there is a fault trend in the power supply circuit and an early warning signal is issued. Polynomial fitting model extrapolation: When a polynomial fitting model is used, each time between the current time and the end time of the preset prediction time window is substituted into the fitted polynomial equation in turn to calculate the predicted value sequence of electrical parameters within the prediction time window; the maximum and minimum values ​​in the predicted value sequence are taken as the judgment criteria. If the extreme value touches the fault judgment boundary, it is determined that the power supply circuit has a fault trend and a warning signal is issued. The upper limit of the rated operating current and the upper and lower limits of the rated operating voltage of the power supply circuit are used as the fault judgment boundary; when the extrapolated predicted value touches the above boundary within the preset prediction time window, it is determined that the power supply circuit has a fault trend. Preferably, the prediction time window can be differentiated. For power supply circuits corresponding to high safety level loads, a relatively longer preset prediction time window (e.g., 100ms) can be used to trigger the pre-switching action with a greater time lead, ensuring uninterrupted power supply to high safety level loads. For power supply circuits corresponding to non-critical loads, a relatively shorter prediction time window (e.g., 20ms) can be used to avoid frequent false triggering of pre-switching actions due to prediction errors.

[0080] Compared with existing passive response switching, predictive early warning based on trend models can advance the system response time window to before the fault is officially triggered, effectively avoiding the risk of power outage caused by fault current surges.

[0081] Preferably, the power supply ratio of the healthy circuit is gradually increased within a preset switching time window (e.g., 5ms) in a gradual manner, while the power supply ratio of the circuit with a fault trend is gradually decreased. The sum of the power supply ratios of the two circuits remains at 100% throughout the entire switching process, ensuring that the load end receives a continuous and uninterrupted power supply.

[0082] This predictive switching mechanism ensures the continuity of power supply to the vehicle's low-voltage loads before a fault actually occurs, which is of great significance for ensuring uninterrupted power supply to high-safety-level loads such as intelligent driving systems, electric power steering systems, and electronic braking systems.

[0083] Example 3 This embodiment provides a vehicle including the aforementioned low-voltage power supply system for autonomous driving, specifically including a high-voltage power battery, a DC-DC converter, a main lithium battery, a redundant lithium battery, an isolation protection module, and a fuse box.

[0084] The high-voltage power battery, as the high-voltage energy storage unit of the vehicle, is electrically connected to the high-voltage input terminal of the DC-DC converter through a high-voltage wiring harness, providing high-voltage DC power input to the DC-DC converter.

[0085] The DC-DC converter converts the high-voltage DC output from the high-voltage power battery into low-voltage DC, and outputs charging current to the main lithium battery and redundant lithium battery respectively through dual isolated output terminals, while simultaneously supplying power to the low-voltage load of the vehicle. The DC-DC converter has dual isolated output terminals, with the first output terminal electrically connected to the main lithium battery and the second output terminal electrically connected to the redundant lithium battery. The two output terminals are electrically isolated from each other, ensuring that a failure in one terminal does not affect the normal operation of the other terminal.

[0086] The main lithium battery has a nominal voltage of 48V and is electrically connected to the first output terminal of the DC-DC converter. It receives the charging current output by the DC-DC converter and is in a charging maintenance state when the high-voltage power system is working normally, while simultaneously discharging power to the low-voltage load of the vehicle. Compared with traditional 12V lithium batteries, the 48V main lithium battery has higher energy density and larger energy storage capacity, providing a longer backup power time for the vehicle after the high-voltage power system fails.

[0087] The redundant lithium battery has a nominal voltage of 12V and is electrically connected to the second output of the DC-DC converter. It receives the charging current from the DC-DC converter and automatically assumes the responsibility of discharging power to the vehicle's low-voltage loads in the event of a failure in the main lithium battery or its power supply link, forming a dual-redundancy protection system. The 12V redundant lithium battery is compatible with the electrical interfaces of existing vehicle low-voltage equipment, facilitating system integration.

[0088] The isolation protection module is integrated inside the DC-DC converter, connected to the DC-DC converter signal, and designed in accordance with ASIL B functional safety level. The isolation protection module has two operating modes: The first is a passive fault isolation mode, used to monitor the voltage, current, and insulation status of each power supply circuit in real time. When any power supply circuit experiences a short circuit, overvoltage, undervoltage, or open circuit anomaly, the faulty circuit is disconnected within milliseconds to achieve electrical isolation, while maintaining continuous power supply to the low-voltage load from the remaining healthy circuits. The second is a predictive fault switching mode, which continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period (e.g., 1ms) to form a time-series sampling sequence. A linear regression or polynomial fitting method is used to establish a trend model of the electrical parameters of each power supply circuit. Based on this trend model, the module extrapolates forward to predict the change trend of electrical parameters of each power supply circuit within a preset prediction time window. When the prediction result indicates that the electrical parameters of any power supply circuit will reach the fault judgment condition within the preset prediction time window, a pre-switching action is performed before the actual fault occurs, smoothly transferring the power supply responsibility to the healthy circuit. For power supply circuits corresponding to high-safety-level loads, a relatively longer preset prediction time window is used to trigger the pre-switching action with a greater time lead, ensuring uninterrupted power supply to high-safety-level loads.

[0089] The fuse box is electrically connected to the main lithium battery and the redundant lithium battery respectively, and is used to distribute the power output from the main lithium battery and the redundant lithium battery to the low-voltage electrical equipment of the vehicle, so as to realize centralized power distribution and overcurrent protection.

[0090] In some implementations, the low-voltage loads of the vehicle include high-safety-level loads and non-critical loads. High-safety-level loads are electrically connected to both the main lithium battery and redundant lithium batteries, employing dual-circuit redundant power supply; non-critical loads are powered by a single circuit. Specifically, high-safety-level loads include at least one of the intelligent driving system, electric power steering system, and electronic braking system. These loads directly relate to the vehicle's ability to safely take over after autonomous driving failure; implementing dual-circuit redundant power supply ensures the vehicle remains basically controllable until the driver takes over. Using a single circuit for non-critical loads effectively reduces the overall system complexity and cost while ensuring the reliability of power supply to high-safety-level loads.

[0091] In some implementations, both the main lithium battery and the redundant lithium battery are installed in the collision safety zone of the passenger compartment, and are located close to the fuse box. Specifically, the collision safety zone is the space under the seats, including under the driver and front passenger seats or under the second-row seats. This space is protected by the vehicle body structure and offers good safety protection in a collision, effectively avoiding mechanical damage to the battery caused by a vehicle collision. Located close to the fuse box, it shortens the power supply harness length, reducing line impedance and the risk of harness failure.

[0092] In some implementations, when a high-voltage power battery or DC-DC converter fails, causing the DC-DC converter to stop working, the main lithium battery and redundant lithium batteries enter an independent power supply mode, providing dual backup power to the vehicle's low-voltage loads. In this mode, even if a single lithium battery fails, the system can still rely on the other battery to maintain power supply to critical loads, significantly improving the system's continuous operating capability under extreme conditions.

[0093] The above system executes the following power supply control process: Step 1: Acquire the operational status information of the autonomous driving system and the working status information of the high-voltage power system. The on-board controller collects the above status information in real time through the communication interface with the autonomous driving domain controller and the high-voltage management system, providing a data foundation for subsequent power supply mode decisions.

[0094] Step 2: Determine the current power supply mode based on the operating status information of the high-voltage power system. When the high-voltage power system is in normal operating condition, control the DC-DC converter as the main power supply to power the low-voltage load of the vehicle, and simultaneously charge the main lithium battery and the redundant lithium battery. At this time, the DC-DC converter continuously replenishes the two sets of lithium batteries through the dual-channel isolated output terminal to keep them in a fully charged standby state. When the high-voltage power system is in a fault state, control the main lithium battery and the redundant lithium battery to enter an independent power supply mode, providing dual-channel backup power to the low-voltage load of the vehicle, thereby responding to extreme conditions such as high-voltage power failure triggered by collision, serious power battery failure, or main relay disconnection. The independent power supply of the dual lithium batteries ensures that the vehicle remains in a basically controllable state.

[0095] Preferably, determining the current power supply mode further includes: such as Figure 2 As shown, when the autonomous driving system fails but the high-voltage power system is operating normally, the DC-DC converter maintains its power supply mode and sends a takeover request to the dynamic driving task support user. Power is continuously supplied to the braking system, steering system, and intelligent driving system until the takeover is completed. This step ensures that in scenarios where the autonomous driving system fails but the high-voltage power system is functioning normally, the system can continuously supply power to safety-critical loads via the DC-DC converter, meeting the mandatory requirements of GB / T 40429-2021 for continuous power supply after L3 level autonomous driving failure.

[0096] Step 3: The isolation protection module continuously monitors the status of each power supply circuit. When any abnormality is detected in any power supply circuit, the abnormal circuit is cut off, while the remaining circuits continue to be powered.

[0097] Specifically, detecting whether an abnormality has occurred in the power supply circuit includes: First, the current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence.

[0098] Secondly, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits based on the time-series sampling sequence. If such anomalies are detected, the isolation protection module is triggered within milliseconds to cut off the faulty branch while maintaining the healthy circuit to continue supplying power to the high-safety-level load, preventing the fault from spreading and ensuring the system's functional safety. If no such anomalies are detected, trend prediction is performed. Then, an electrical parameter trend model is established for the time-series sampling sequence. Based on the trend model, the electrical parameter change trend of each power supply circuit within a preset prediction time window is predicted. When the prediction result shows that the electrical parameter of any power supply circuit will reach the abnormal judgment condition within the preset prediction time window, the isolation protection module is triggered in advance to perform a pre-switching action on the circuit before the actual occurrence of the fault, so as to smoothly transfer the power supply responsibility of the circuit to the healthy circuit.

[0099] Furthermore, based on the time-series sampling sequence, the system detects whether there are short circuits, overvoltages, undervoltages, or open circuits. Specifically, it compares the current and voltage values ​​of adjacent sampling points in the time-series sampling sequence in real time and identifies four types of anomalies based on the following characteristics: (1) Short circuit anomaly: The current value suddenly increases between adjacent sampling points and the voltage value drops in the same step, indicating that a low impedance short circuit fault has occurred in the power supply circuit; preferably, within a single sampling period, the increase in the current value exceeds a preset multiple of the rated operating current, which is set to 3 times in this embodiment, and is thus judged as a sudden increase in current; at the same time, within the same sampling period, the drop in the voltage value exceeds a preset proportion of the rated operating voltage, which is taken as 20%, and the short circuit anomaly is judged when both conditions are met.

[0100] (2) Overvoltage abnormality: The voltage value continuously exceeds the upper limit of the normal working range, indicating that the power supply circuit is overcharged or the voltage regulation fails; preferably, in this embodiment, the voltage values ​​of three or more consecutive sampling points continuously exceed the upper limit of the normal working range.

[0101] (3) Undervoltage abnormality: The voltage value is consistently lower than the lower limit of the normal operating range, indicating that the battery capacity is insufficient or there is a high impedance fault in the power supply link.

[0102] (4) Open circuit anomaly: The current value drops sharply to near zero and the voltage value rises abnormally, indicating that an open circuit fault has occurred in the power supply circuit. Specifically, if the current value drops by more than a preset proportion of the rated operating current within a single sampling period (set to 90%) and the current value after the drop is lower than the preset lower limit of the rated operating current (set to 5%), it is judged as a sharp drop in current to near zero; if the voltage value increases by more than a preset proportion of the rated operating voltage within the same sampling period (set to 15%), it is judged as an abnormal voltage rise; an open circuit anomaly is judged only when both conditions are met simultaneously.

[0103] When any of the above features are detected, it is determined that a corresponding abnormality has occurred in the power supply circuit. Within milliseconds, the isolation protection module is triggered to disconnect the faulty branch, while the healthy circuit continues to supply power to the high-safety-level load, preventing the fault from spreading and ensuring the functional safety of the system.

[0104] It should be understood that the normal operating range is calibrated during system initialization based on the battery rating parameters and the vehicle's low-voltage system specifications, or pre-configured at the factory.

[0105] Furthermore, predictive failover includes the following steps: (1) Trend modeling: The isolation protection module continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period to form a time-series sampling sequence of each power supply circuit; the time-series sampling sequence is used to establish an electrical parameter trend model of each power supply circuit by using linear regression or polynomial fitting method, so as to provide a data basis for fault prediction.

[0106] Specifically, linear regression modeling: using the timestamp of each sampling point in the time-series sampling sequence as the independent variable and the corresponding current value or voltage value as the dependent variable, the least squares method is used to fit a linear equation to obtain the slope of the electrical parameters changing with time; the sign of the slope reflects the direction of parameter change, and the absolute value of the slope reflects the rate of parameter degradation. When the absolute value of the slope continues to increase, it indicates that there is an accelerated degradation trend in the power supply circuit.

[0107] Polynomial fitting modeling: When electrical parameters exhibit a nonlinear degradation trend, a quadratic or cubic polynomial is used to fit the time-series sampling sequence to more accurately capture the curvature characteristics of accelerated degradation. The quadratic polynomial is suitable for uniformly accelerated degradation scenarios, while the cubic polynomial is suitable for complex scenarios where the degradation rate itself is also changing.

[0108] Furthermore, the model selection strategy is as follows: linear regression or polynomial fitting is adaptively selected based on the root mean square value (RMSE) of the fitting residuals of the time-series sampling sequence. When the RMSE of linear regression is lower than the preset residual threshold, the linear regression model is used; otherwise, the polynomial fitting model is switched to achieve a balance between computational cost and fitting accuracy.

[0109] In this embodiment, the preset residual threshold can be calibrated according to the actual parameters of the vehicle's low-voltage system. For example, for current signals, the preset residual threshold can be set to 0.5A; for voltage signals, the preset residual threshold can be set to 0.1V. When the RMSE of the linear regression model is lower than the above-mentioned preset residual threshold, the linear regression model is used; otherwise, the model is switched to a multinomial fitting model.

[0110] The above modeling process is updated on a rolling basis in each sampling period. The trend model is updated once a new sampling point is acquired, ensuring that the trend model always reflects the latest state of the power supply circuit and provides a real-time data basis for fault prediction.

[0111] (2) Trend prediction and early warning: Based on the trend model, the electrical parameter change trend of each power supply circuit is predicted within the preset prediction time window; when the prediction result shows that the electrical parameter of any power supply circuit will reach the fault judgment condition within the preset prediction time window, the system determines that the power supply circuit has a fault trend and issues an early warning signal.

[0112] The specific process is as follows: Linear regression model extrapolation: Substitute the slope of the trend model of linear regression and the measured values ​​of electrical parameters at the current moment into the linear equation to calculate the predicted value at the end of the preset prediction time window; if the predicted value exceeds the fault judgment range, it is determined that there is a fault trend in the power supply circuit and an early warning signal is issued. Polynomial fitting model extrapolation: When a polynomial fitting model is used, each time between the current time and the end time of the preset prediction time window is substituted into the fitted polynomial equation in turn to calculate the predicted value sequence of electrical parameters within the prediction time window; the maximum and minimum values ​​in the predicted value sequence are taken as the judgment criteria. If the extreme value touches the fault judgment boundary, it is determined that the power supply circuit has a fault trend and a warning signal is issued. The upper limit of the rated operating current and the upper and lower limits of the rated operating voltage of the power supply circuit are used as the fault judgment boundary; when the extrapolated predicted value touches the above boundary within the preset prediction time window, it is determined that the power supply circuit has a fault trend. Preferably, the prediction time window can be differentiated. For power supply circuits corresponding to high safety level loads, a relatively longer preset prediction time window (e.g., 100ms) can be used to trigger the pre-switching action with a greater time lead, ensuring uninterrupted power supply to high safety level loads. For power supply circuits corresponding to non-critical loads, a relatively shorter prediction time window (e.g., 20ms) can be used to avoid frequent false triggering of pre-switching actions due to prediction errors.

[0113] Compared with existing passive response switching, predictive early warning based on trend models can advance the system response time window to before the fault is officially triggered, effectively avoiding the risk of power outage caused by fault current surges.

[0114] Preferably, the power supply ratio of the healthy circuit is gradually increased within a preset switching time window (e.g., 5ms) in a gradual manner, while the power supply ratio of the circuit with a fault trend is gradually decreased. The sum of the power supply ratios of the two circuits remains at 100% throughout the entire switching process, ensuring that the load end receives a continuous and uninterrupted power supply.

[0115] This predictive switching mechanism ensures the continuity of power supply to the vehicle's low-voltage loads before a fault actually occurs, which is of great significance for ensuring uninterrupted power supply to high-safety-level loads such as intelligent driving systems, electric power steering systems, and electronic braking systems.

[0116] Example 4 This embodiment provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the aforementioned low-voltage power supply method for autonomous driving.

[0117] Example 5 This embodiment provides a computer program product including executable instructions, which are computer instructions; the executable instructions are stored in a computer-readable storage medium. When the vehicle's processor reads the executable instructions from the computer-readable storage medium and executes the executable instructions, the steps of the aforementioned low-voltage power supply method for autonomous driving are completed.

[0118] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0119] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0120] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment, whereby a series of operational steps are performed to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0121] The steps and methods involved in Embodiments 2 to 5 above correspond to those in Embodiment 1. For detailed implementation methods, please refer to the relevant description section of Embodiment 1. The descriptions of each embodiment in the above embodiments have different emphases. Parts not described in detail in a certain embodiment can be referred to the relevant descriptions of other embodiments.

[0122] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0123] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A low-voltage power supply method for autonomous driving, characterized in that, include: Obtain the operating status information of the high-voltage power system; The current power supply mode is determined based on the operating status information of the high-voltage power system; When the high-voltage power system is in normal working condition, the DC-DC converter is controlled to supply power to the low-voltage load of the vehicle as the main power supply, and at the same time, the main lithium battery and the redundant lithium battery are charged; when the high-voltage power system is in fault condition, the main lithium battery and the redundant lithium battery are controlled to enter independent power supply mode, providing dual backup power to the low-voltage load of the vehicle respectively. The isolation protection module continuously monitors the status of each power supply circuit. When any abnormality is detected in any power supply circuit, the abnormal circuit is cut off, while the remaining circuits continue to be powered.

2. The low-voltage power supply method for autonomous driving as described in claim 1, characterized in that, The method further includes: when the autonomous driving system fails and the high-voltage power system is in normal working condition, maintaining the DC-DC converter power supply mode, issuing a takeover request to the dynamic driving task support user, and continuously supplying power to the braking system, steering system and intelligent driving system until the takeover is completed.

3. The low-voltage power supply method for autonomous driving as described in claim 1, characterized in that, The above-mentioned detection of the status of each power supply circuit, when an abnormality is detected in any power supply circuit, disconnects the abnormal circuit while maintaining continuous power supply to the remaining circuits, specifically including: The current and voltage values ​​of each power supply circuit are continuously collected at a preset sampling period to form a time-series sampling sequence; The system detects whether there are any abnormalities such as short circuit, overvoltage, undervoltage, or open circuit based on the time-series sampling sequence; if the above abnormalities are detected, the isolation protection module is triggered to cut off the faulty branch. If no abnormality is detected, trend prediction is performed; an electrical parameter trend model is established for the time-series sampling sequence, and the electrical parameter change trend of each power supply circuit is predicted based on the trend model within a preset prediction time window. When the prediction result indicates that the electrical parameter of any power supply circuit will reach the abnormal judgment condition within the preset prediction time window, the isolation protection module is triggered in advance to perform a pre-switching action on the circuit before the actual occurrence of the fault, so as to smoothly transfer the power supply responsibility of the circuit to the healthy circuit.

4. A low-voltage power supply method for autonomous driving as described in claim 3, characterized in that, The establishment of an electrical parameter trend model for the time-series sampling sequence specifically involves: The isolation protection module continuously collects the current and voltage values ​​of each power supply circuit at a preset sampling period to form a time-series sampling sequence for each power supply circuit; the time-series sampling sequence is used to establish an electrical parameter trend model for each power supply circuit, including a linear regression model and a multinomial fitting model. The system adaptively selects between linear regression and polynomial fitting based on the root mean square value of the fitting residuals of the time-series sampling sequence. When the root mean square value of linear regression is lower than the preset residual threshold, the linear regression model is used; otherwise, the system switches to the polynomial fitting model.

5. A low-voltage power supply method for autonomous driving as described in claim 3, characterized in that, The trend model prediction includes: Linear regression model extrapolation: Substitute the slope of the trend model of linear regression and the measured values ​​of electrical parameters at the current moment into the linear equation to calculate the predicted value at the end of the preset prediction time window; if the predicted value exceeds the fault judgment range, it is determined that the power supply circuit has a fault trend. Polynomial fitting model extrapolation: When a polynomial fitting model is used, each time between the current time and the end time of the preset prediction time window is substituted into the fitted polynomial equation in turn to calculate the predicted value sequence of electrical parameters within the prediction time window; the maximum and minimum values ​​in the predicted value sequence are taken as the judgment criteria. If the extreme value touches the fault judgment boundary, it is determined that the power supply circuit has a fault trend.

6. A low-voltage power supply method for autonomous driving as described in claim 1, characterized in that, The low-voltage load of the vehicle includes high-safety-level loads and non-critical loads. The high-safety-level loads are electrically connected to both the main lithium battery and the redundant lithium battery, and are powered by dual-circuit redundant power supply. The non-critical loads are powered by a single circuit. The high-safety-level load includes at least one of an intelligent driving system, an electric power steering system, and an electronic braking system.

7. A low-voltage power supply system for autonomous driving, characterized in that, It includes a high-voltage power battery and a DC-DC converter; a main lithium battery, electrically connected to the DC-DC converter, for receiving the charging current output by the DC-DC converter and discharging power to the low-voltage load of the vehicle; a redundant lithium battery, electrically connected to the DC-DC converter, for receiving the charging current output by the DC-DC converter and discharging power to the low-voltage load of the vehicle when the main lithium battery or its power supply link fails; and an isolation protection module, integrated inside the DC-DC converter and signal-connected to the DC-DC converter, for real-time detection of the status of each power supply circuit, cutting off the faulty circuit when any power supply circuit is detected as abnormal, and keeping the remaining circuits supplying power to the low-voltage load.

8. A vehicle, characterized in that, Includes the low-voltage power supply system for autonomous driving as described in claim 7.

9. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, complete the low-voltage power supply method for autonomous driving as described in any one of claims 1-6.

10. A computer program product comprising executable instructions stored in a computer-readable storage medium; wherein, When the vehicle's processor reads executable instructions from a computer-readable storage medium and executes the executable instructions, it completes the low-voltage power supply method for autonomous driving as described in any one of claims 1-6.

Images