An EOL testing method and system for an automobile intelligent electric appliance box

Abstract

The application discloses an EOL testing method and system for an intelligent electric appliance box of an automobile, and belongs to the technical field of automobile electronic testing. The method comprises the following steps: deploying distributed synchronous sampling nodes, and synchronously sampling each output channel at the same time through high-precision clocks; after sampling data is read and packaged by a data transfer unit, the sampling data is sent to an upper computer through a bus; during the testing process, the sampling range is automatically switched according to the test voltage and / or the channel current size, and the optimal range is matched; the abnormal mode of the sampling link is identified based on a CNN model, and the sampling link is compensated and calibrated according to the identification result, so as to correct the sampling data. Through the distributed synchronous sampling, the intelligent range automatic matching and the CNN self-calibration compensation, the application realizes the multi-channel high-precision synchronous testing, reduces the testing cost and the manual calibration frequency, and is suitable for the batch offline detection of the intelligent electric appliance box.

Description

Technical Field

[0001] This invention relates to the field of automotive electronics testing technology, and in particular to an EOL testing method and system for automotive smart electrical boxes. Background Technology

[0002] With the development of vehicle electrification and intelligence, the future vehicle power distribution architecture places higher demands on safety, reliability, and flexibility. Traditional power distribution architectures mainly rely on mechanical fuses and relays, which are simple in structure and low in cost, but have drawbacks such as non-resettable after a fuse blows, slow response, high susceptibility to environmental influences, and lack of real-time diagnostics. In contrast, intelligent power distribution architectures use semiconductor devices (such as electronic fuses, high-side switches, and MOSFETs) to achieve rapid isolation, precise control, real-time monitoring, and self-recovery functions, meeting the requirements of autonomous driving (L3-L5) and drive-by-wire systems for "failure-safe operation" and "rapid fault isolation" in the power distribution network.

[0003] The intelligent power distribution architecture is shifting from centralized to distributed, reducing wiring harnesses and improving reliability through area control. Electronic fuses, as a core component, integrate protection and switching, supporting I / O... 2 Features include T-harness protection, OTA upgrades, and low-power constant power supply. Specific application scenarios include: load control (LCSP), load power supply protection (LSP), power connection and isolation (PCI), power protection (BP), data exchange (DX), harness protection (WP), power isolation (PSP), parking constant power distribution (ADP), and voltage conversion isolation (VCI). Rapid response is particularly crucial when dealing with voltage dips caused by short circuits.

[0004] As a core component of intelligent power distribution architecture, the intelligent electrical box requires voltage and load current measurements on each output channel during its End of Line (EOL) phase to verify its load-bearing capacity and protection functions. Currently, sampling accuracy is limited by noise, temperature drift, and ADC resolution; multi-channel polling sampling is time-consuming and cannot respond to transient faults. Traditional sampling architectures struggle to simultaneously meet the industry demands of "low cost + high precision + high speed + large number of channels." A detailed comparison between traditional and intelligent power distribution architectures is shown in the table below: Summary of the Invention

[0005] This invention is made to solve the above-mentioned problems, and its purpose is to provide an EOL testing method and system for automotive smart electrical boxes.

[0006] This invention provides an EOL (End of Operation) test method for an automotive smart electrical box, comprising the following steps: Step 1, deploying distributed synchronous sampling nodes: Multiple sampling nodes are deployed in each output channel area of ​​the smart electrical box under test. Each sampling node is synchronized by a high-precision clock and independently and synchronously samples the voltage and current of all output channels at the same time to obtain the voltage and current sampling values ​​of each channel; Step 2, data packaging and forwarding: Each sampling node sends the sampled data to a data relay unit. The data relay unit reads the sampled data, packages the data, and sends it to the host computer via a bus; Step 3, intelligent range automatic matching: During the test, the sampling range is automatically switched according to the current test voltage and / or the current magnitude of each channel to match the optimal range of the sampling signal of each channel; Step 4, self-calibration compensation: Abnormal patterns of the sampling link are identified based on a CNN model. The sampling link is compensated and calibrated according to the identification results to correct the sampling data and output the calibrated test results.

[0007] Preferably, in step 1, the sampling node uses a 16-bit, 8-channel synchronous sampling ADC chip, and each sampling node samples 8 output channels simultaneously. Multiple sampling nodes work together to achieve synchronous sampling of all output channels.

[0008] Preferably, in step 2, the data relay unit adopts an "interrupt + ping-pong buffer" architecture, using two buffers to alternately receive sampled data. When one buffer is full, data transmission is triggered, and data continues to be filled into the other buffer to ensure that the data flow is not interrupted.

[0009] Preferably, in step 3, the test voltage includes three voltage states: 8V, 14V, and 16V, corresponding to undervoltage, standard, and overvoltage states, respectively; the automatic switching of the sampling range includes: using a first voltage division ratio in the 8V test scenario, and using a second voltage division ratio in the 14V and 16V test scenarios, wherein the first voltage division ratio is greater than the second voltage division ratio.

[0010] Preferably, in step 3, the range switching is achieved by voltage division switching through an optocoupler relay.

[0011] Preferably, in step 4, the abnormal modes include one or more of the following: power supply output voltage deviation, probe contact voltage drop deviation, line voltage drop deviation caused by heat, relay contact voltage drop deviation, and load resistance change due to heat. The CNN model takes the voltage waveform and current waveform within a preset time window as input, compares the correlation of the sampled values ​​of each channel, outputs the abnormal mode classification result, and reports the abnormal sampling channel. The CNN model includes residual connections and a global attention mechanism.

[0012] Preferably, the self-calibration compensation is performed automatically when the test system is idle, without the need for external calibration during production shutdown.

[0013] This invention provides an EOL (End-of-Life) testing system for an automotive smart electrical box, comprising: a distributed sampling module including multiple sampling nodes, each sampling node being synchronized via a high-precision clock, used to synchronously sample each output channel of the smart electrical box under test at the same time, obtaining voltage and current sampling values ​​for each channel; a data relay module connected to the distributed sampling module, used to read the sampled data, package the data, and forward it; and a host computer connected to the data relay module, used to receive the sampled data, perform range switching, perform CNN model inference and self-calibration compensation, and output the calibrated test results.

[0014] The present invention provides an electronic device, including a memory and a processor. The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the method described above.

[0015] The present invention provides a computer storage medium storing computer-executable instructions thereon, which, when executed by a processor, implement the steps of the method described above.

[0016] Technical effect Compared with existing technologies, the EOL testing method and system for automotive intelligent electrical boxes provided by this invention, with electronic fuses as the core and combined with distributed area control, achieves fast, accurate, diagnosable, and resettable power distribution management, and meets the functional safety requirements of standards such as VDA 450. It includes two sets of power grid safety redundancy designs (A and B), comprising two main modules: basic vehicle safety load and intelligent driving system safety load, with power distribution management, diagnosis, and communication for a total of 70 loads. Specifically, it has the following technical effects: Through a distributed synchronous sampling network, each node independently completes analog-to-digital conversion, and high-precision clock synchronization enables coordinated sampling of all channels at the same time point. Compared with traditional centralized polling sampling, this scheme achieves independent high-precision sampling for each channel, significantly improving the accuracy of multi-channel synchronous testing and effectively shortening the total testing time. It can perfectly capture the true phase relationship of voltage / current in each channel.

[0017] By automatically matching the intelligent range, the sampling range is dynamically switched according to the magnitude of the test voltage and / or current, ensuring that small current signals are not drowned out by noise and large current signals do not saturate the ADC, thus achieving high-precision current detection with a wide range from 1A to 200A.

[0018] By identifying abnormal patterns in the sampling link using a CNN model, the EOL testing system can automatically perform self-calibration of the sampling link during idle periods. A fault diagnosis algorithm is designed to proactively detect and report sampling channel anomalies by comparing the correlation of sampled values ​​from different channels. This method is significant for ensuring the quality of long-term operational testing and reducing the cost of manual calibration. Attached Figure Description

[0019] Figure 1 This is an overall architecture diagram of the smart appliance box EOL testing system provided in this embodiment of the invention; Figure 2 This is a schematic diagram of the physical structure of the smart appliance box product in an embodiment of the present invention; Figure 3 This is a comparison chart of the response speeds of traditional power distribution architecture and intelligent power distribution architecture in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the deployment and load connection of distributed synchronous sampling nodes in an embodiment of the present invention; Figure 5 This is a schematic diagram of the signal acquisition and data forwarding hierarchy in an embodiment of the present invention; Figure 6 This is a physical application diagram of the EOL testing system in an embodiment of the present invention. Detailed Implementation

[0020] To make the technical means, creative features, objectives and effects of this invention easy to understand, the following embodiments, in conjunction with the accompanying drawings, specifically illustrate an EOL testing method and system for an automotive intelligent electrical box according to this invention.

[0021] This embodiment provides an EOL (End of Operation) test method for automotive smart electrical boxes.

[0022] Figure 1 This is the overall architecture diagram of the EOL testing system for the smart appliance box provided in this embodiment.

[0023] like Figure 1 As shown, the EOL testing system of this embodiment includes a programmable power supply, an EOL testing vehicle, a load cell, and a distributed sampling network. Based on the above system architecture, this embodiment provides an EOL testing method for a smart electrical box.

[0024] Figure 2 This is a schematic diagram of the physical structure of the smart appliance box product in this embodiment.

[0025] like Figure 2 As shown, the test object in this embodiment is: intelligent electrical box (IDU output distribution), which includes OT terminal and three connectors J1, J2, J3, with a total of 8 channels of 40A, 10 channels of 20A, 10 channels of 10A, 6 channels of 4A, 14 channels of 2.5A, and 17 channels of 1A rated load output.

[0026] The EOL testing method for the automotive intelligent electrical box provided in this embodiment includes the following steps: Step S1, Deploy distributed synchronous sampling nodes: Deploy multiple sampling nodes in each output channel area of ​​the smart electrical box under test. Each sampling node is synchronized by a high-precision clock and independently and synchronously samples the voltage and current of all output channels at the same time to obtain the voltage and current sampling values ​​of each channel.

[0027] Figure 3 This is a comparison chart of the response speeds of the traditional power distribution architecture and the intelligent power distribution architecture in this embodiment.

[0028] like Figure 3 As shown, there are significant differences between the traditional power distribution architecture and the intelligent power distribution architecture used in this embodiment in terms of response speed, resetability, and diagnostic functions. This embodiment deploys multiple sampling nodes in each output channel area of ​​the intelligent electrical appliance box under test.

[0029] Figure 4 This is a schematic diagram of the deployment and load connection of distributed synchronous sampling nodes in this embodiment.

[0030] like Figure 4 As shown, in this embodiment, each sampling node uses a 16-bit, 8-channel synchronous sampling ADC chip, and each sampling node samples 8 output channels simultaneously. Multiple sampling nodes are synchronized by a high-precision clock, independently and synchronously sampling the voltage and current of all output channels at the same time to obtain the voltage and current sample values ​​of each channel. Each node independently completes analog-to-digital conversion, and the high-precision clock synchronization enables coordinated sampling of all channels at the same time point, thereby perfectly capturing the true phase relationship of voltage / current of each channel.

[0031] In this embodiment, the CM2248, a 16-bit, 8-channel synchronous sampling ADC chip from Beijing Shimo Microelectronics, is used, primarily for high-precision data acquisition scenarios. Each channel has a maximum throughput of 200kSPS and a signal-to-noise ratio of 92dB. Input range: Supports ±10V or ±5V bipolar input, controlled by the RANGE pin; analog input impedance is 1MΩ, requiring no front-end driver amplifier. Power supply and power consumption: Single 5V power supply; operating power consumption is approximately 107.5mW; standby power consumption is approximately 28.5mW; shutdown mode power consumption can be as low as 5μW.

[0032] Step S2, Data Packaging and Forwarding: Each sampling node sends the sampled data to the data transfer unit. The data transfer unit reads the sampled data, packages the data, and sends it to the host computer via the bus.

[0033] Figure 5 This is a schematic diagram of the signal acquisition and data forwarding hierarchy in this embodiment.

[0034] like Figure 5As shown, each sampling node sends its sampled data to the data relay unit. The data relay unit uses a C8051F040 microcontroller, which is the "central hub" of the entire data acquisition and forwarding chain. After reading the data from the CM2248, it repackages the data and sends it to the host computer via a CAN transceiver and CAN bus. The data relay unit employs an "interrupt + ping-pong buffer" software architecture, meaning it uses two buffers to alternately receive ADC data. When one buffer is full, a CAN transmission is triggered, and the program continues to fill the other buffer with data, ensuring an uninterrupted data flow.

[0035] Step S3, Intelligent Range Automatic Matching: During the test, the sampling range is automatically switched according to the current test voltage and / or the current of each channel, so that the sampling signal of each channel matches the optimal range.

[0036] In the EOL (End-of-Life) test of automotive intelligent electrical boxes, three voltages—8V, 14V, and 16V—are required. The core purpose is to simulate and verify the three most typical voltage states in automotive electrical systems, ensuring stable operation under any real-world conditions. These three voltage values ​​represent: ① 8V - Simulate "undervoltage": Simulates the low-voltage state when the battery is severely depleted or the engine is cold-started.

[0037] ② 14V - Simulated "Standard": Simulates the rated voltage when the engine is working normally and the generator is charging the battery.

[0038] ③ 16V - Simulate "overvoltage": Simulates the high voltage output under extreme conditions such as generator regulator failure.

[0039] The customized rated load sizes under 14V conditions are: 2 DC-DC converters each with a 200A range, 8 channels with a 40A range, 10 channels with a 20A range, 10 channels with a 10A range, 6 channels with a 4A range, 14 channels with a 2.5A range, and 17 channels with a 1A range. When switching between 8V and 16V testing, the current magnitude of each channel will change accordingly. The load current of each output channel of the device under test varies significantly. A fixed sampling range will cause small current signals to be submerged in local noise, while large currents will saturate the ADC. In this embodiment, the CM2248 is used for gain amplification programmable adjustment. The sampling module pre-detects the voltage range during the test initialization phase, automatically matches the optimal range, and dynamically tracks current changes to switch the range during the test. A low-cost domestic optocoupler relay APW214 is used for voltage divider range switching. The first voltage division ratio (one-quarter voltage division in this embodiment) corresponds to the 8V test scenario, and the second voltage division ratio (one-eighth voltage division in this embodiment) corresponds to the 14V and 16V test scenarios. The low-side output voltage is not divided and the gain is adjusted using CM2248.

[0040] Step S4, self-calibration compensation: Based on the CNN model, abnormal patterns in the sampling link are identified, and the sampling link is compensated and calibrated according to the identification results to correct the sampling data and output the calibrated test results.

[0041] For end-of-life (EOL) testing of smart electrical boxes, the complexity of the entire system and the high requirements of the testing conditions inevitably introduce significant instability factors. These include deviations in the power supply output voltage, voltage drop deviations at each contact between the product and the probe, deviations in line voltage drop after the probe and wire heat up during continuous testing, deviations in contact voltage drop of circuit breakers and relays controlling load conduction, deviations in the impact of resistance changes caused by continuous load operation and heat generation on the overall circuit impedance, and zero-point and gain drift of the sampling sensor after long-term operation.

[0042] Traditionally, external calibration is performed by halting production, which is costly. This embodiment addresses the aforementioned deviations by building a model based on a CNN algorithm, enabling the EOL testing system to automatically perform self-calibration of the sampling link during idle periods. A fault diagnosis algorithm is designed to proactively detect and report sampling channel anomalies by comparing the correlation of sampled values ​​from different channels. This method is significant for ensuring the quality of long-term operational testing and reducing the cost of manual calibration.

[0043] In this embodiment, anomaly patterns in the sampling link are identified based on a CNN model. The sampling link is then compensated and calibrated according to the identification results to correct the sampling data, and the calibrated test results are output. Specifically, the CNN model includes residual connections and a global attention mechanism to improve sensitivity to different anomalies. The CNN model takes voltage and current waveforms within a preset time window as input, compares the correlation of sampled values ​​from each channel, outputs anomaly pattern classification results, and reports sampling channel anomalies. Details are as follows: This embodiment also provides an EOL testing system for automotive intelligent electrical boxes. It includes a distributed sampling module, a data relay module, and a host computer. The system employs a CM2248 at the bottom layer for high-precision synchronous data acquisition, a C8051F040 in the middle for data packaging and forwarding, and the host computer for core data processing, model inference, and human-computer interaction.

[0044] The distributed sampling module includes multiple sampling nodes, which are synchronized by a high-precision clock to simultaneously sample each output channel of the smart appliance box under test, obtaining the voltage and current sampling values ​​of each channel.

[0045] The data relay module is connected to the distributed sampling module and is used to read the sampled data, package the data, and forward it.

[0046] The host computer is connected to the data transfer module to receive sampled data, perform range switching, execute CNN model inference and self-calibration compensation, and output the calibrated test results.

[0047] The working principles and technical details of the distributed sampling module, data transfer module, and host computer correspond to the method steps in this embodiment, and will not be repeated here.

[0048] Figure 6 This is a physical application diagram of the EOL testing system in an embodiment of the present invention.

[0049] The above embodiments are preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention.

[0050] Comparative Example To verify the technical effectiveness of this invention, the EOL testing system of this invention is compared with two typical testing schemes in the prior art. Scheme 1 uses CM2248+C8051F040 (CAN) provided in the above embodiment; Scheme 2 uses AD620+C8051F580 (CAN); and Scheme 3 uses AD620+signal adapter board+NI PCI-6225 (PCI bus). The comparison results are shown in the table below: As can be seen from the table above, apart from the second point (automatic range matching) and the third point (CNN algorithm self-calibration), the advantages of this invention's main synchronous sampling circuit compared to two other existing solutions on the market are obvious in terms of innovation, test synchronization accuracy, manufacturing cost, manufacturing cycle, and ease of maintenance. It can be applied to the EOL mass production of smart appliance boxes; the equipment is accurate and stable, and can be summarized into a replicable template for horizontal application in many other projects.

Claims

1. An EOL (End of Service) test method for an automotive intelligent electrical box, characterized in that, Includes the following steps: Step 1, Deploy distributed synchronous sampling nodes: Deploy multiple sampling nodes in each output channel area of ​​the smart appliance box under test. Each sampling node is synchronized by a high-precision clock and independently and synchronously samples the voltage and current of all output channels at the same time to obtain the voltage and current sampling values ​​of each channel. Step 2, Data Packaging and Forwarding: Each sampling node sends the sampling data to the data relay unit, which reads the sampling data, packages the data, and sends it to the host computer via the bus; Step 3, Intelligent Range Automatic Matching: During the test, the sampling range is automatically switched according to the current test voltage and / or the current of each channel, so that the sampling signal of each channel matches the optimal range; Step 4, self-calibration compensation: Based on the CNN model, abnormal patterns in the sampling link are identified, and the sampling link is compensated and calibrated according to the identification results to correct the sampling data and output the calibrated test results.

2. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: In step 1, the sampling node uses a 16-bit, 8-channel synchronous sampling ADC chip. Each sampling node samples 8 output channels simultaneously, and multiple sampling nodes work together to achieve synchronous sampling of all output channels.

3. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: In step 2, the data relay unit adopts an "interrupt + ping-pong buffer" architecture, using two buffers to alternately receive the sampled data. When one buffer is full, data transmission is triggered, and data continues to be filled into the other buffer to ensure that the data flow is not interrupted.

4. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: In step 3, the test voltage includes three voltage states: 8V, 14V, and 16V, which correspond to undervoltage, standard, and overvoltage states, respectively. The automatic switching of sampling range includes: using a first voltage division ratio in the 8V test scenario, and using a second voltage division ratio in the 14V and 16V test scenarios, wherein the first voltage division ratio is greater than the second voltage division ratio.

5. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: In step 3, the range switching is achieved by voltage division switching through an optocoupler relay.

6. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: In step 4, the abnormal mode includes one or more of the following: power supply output voltage deviation, probe contact voltage drop deviation, line voltage drop deviation caused by heat, relay contact voltage drop deviation, and load resistance change due to heat. The CNN model takes voltage and current waveforms within a preset time window as input, compares the correlation of sampled values ​​of each channel, outputs abnormal pattern classification results, and reports abnormal sampling channels. The CNN model includes residual connections and a global attention mechanism.

7. The EOL test method for automotive intelligent electrical box according to claim 1, characterized in that: The self-calibration compensation is performed automatically when the test system is idle, without the need for external calibration during production shutdown.

8. An EOL testing system for an automotive intelligent electrical box, characterized in that, include: The distributed sampling module includes multiple sampling nodes, which are synchronized by a high-precision clock to simultaneously sample each output channel of the smart electrical box under test, thereby obtaining the voltage and current sampling values ​​of each channel. The data relay module, connected to the distributed sampling module, is used to read the sampled data, package the data, and forward it. The host computer is connected to the data transfer module and is used to receive sampled data, perform range switching, perform CNN model inference and self-calibration compensation, and output the calibrated test results.

9. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the computer-executable instructions are executed by a processor, they implement the steps of the method as described in any one of claims 1-7.

Images