An ai robot multi-loop esd dynamic shielding protection system with built-in self-test

Abstract

The application discloses an AI robot multi-loop ESD dynamic shielding protection system with self-checking function, which comprises a grounding reference module, a layered multi-loop static electricity collection array, a loop static electricity coupling impedance AI adaptive identification unit, a self-checking shielding integrated closed-loop regulation and control unit, a hierarchical dynamic shielding execution module and a local cloud linkage fault research and judgment storage module. The AI robot multi-loop ESD dynamic shielding protection system with self-checking function is based on coupling impedance conjugate matching dynamic shielding, and the multi-loop cross-loop static electricity mutual coupling suppression rate is increased by 51%. The nanosecond-level ESD static electricity impact discharge speed is increased by 63%. Under high and low temperature dust complex industrial working conditions, the robot communication packet loss rate and sensing failure probability are reduced by 98%. The AI robot is suitable for heavy load, high speed and outdoor full-category AI robots.

Description

Technical Field

[0001] This invention relates to the field of robot-related products, specifically to a multi-loop ESD dynamic shielding and protection system for AI robots equipped with fault self-diagnosis. Background Technology

[0002] AI robots integrate multiple electrical circuits, including servo power supply circuits, Ethernet communication circuits, visual sensing signal circuits, and encoder servo feedback circuits. The high-density wiring and close-coupled arrangement of these multiple circuits make them highly susceptible to electrostatic coupling caused by triboelectric charging, equipment arcing, and alternating electric fields in industrial workshops. This can generate nanosecond-level ESD electrostatic discharge shocks. Mild damage can cause noise in the robot's visual images, packet loss on the communication bus, and servo positioning deviations. Severe damage can break down circuit MOSFETs and sensor chips, causing the entire robot to stop and be scrapped. Therefore, a dedicated ESD shielding protection system is an essential electromagnetic compatibility accessory for AI robots.

[0003] Currently, AI robot ESD protection technology has four major shortcomings:

[0004] 1. Static and fixed shielding mode with extremely poor adaptability: Existing protection systems use shielding layers of fixed thickness and electrostatic discharge branches with fixed resistance values, which are only suitable for standard no-load working conditions of robots. After changes in robot load, joint torsion, and changes in ambient temperature and humidity, the electrostatic coupling impedance between the loop and ground and between loops changes dynamically in real time. Fixed shielding parameters cannot match the dynamic coupling impedance, which easily leads to two failure problems: shielding oversaturation and shielding undersaturation. Cross-loop electrostatic crosstalk cannot be eliminated.

[0005] 2. Separate architecture for self-testing and shielding, resulting in resource redundancy and delayed linkage: On the market, fault self-testing modules and ESD shielding modules are independent hardware architectures, each configured with sampling computing power and control chips, resulting in high hardware costs and large space occupation for robot wiring; moreover, self-testing can only detect the continuity of shielding circuits and cannot rely on electrostatic coupling characteristics to determine implicit insulation aging or grounding faults. After the self-test is completed, shielding control commands need to be issued after a delay. Under millisecond-level ESD transient impacts, the control is delayed, resulting in a high probability of protection failure.

[0006] 3. Undifferentiated management of multiple circuits exacerbates coupling crosstalk: The existing all-domain integrated shielding structure uses the same shielding strength for power circuits and sensing circuits. The weak sensing circuits are easily affected by secondary electrostatic interference coupled from the power circuits. The mutual coupling impedance of multiple circuits cannot be decoupled and managed, and the electrostatic accumulation exceeds the standard during long-term operation.

[0007] 4. Weak fault tracing capabilities and passive operation and maintenance: Traditional protection can only alarm whether there is an electrostatic fault, and cannot distinguish between plate aging, cable damage, grounding fault, and external ESD impact fault types. There is no impedance tracing data, and the only option is to disassemble the machine and repair it point by point, resulting in high robot downtime and maintenance costs. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis, so as to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: an AI robot multi-loop ESD dynamic shielding protection system equipped with fault self-diagnosis, comprising a grounding reference module, a layered multi-loop electrostatic acquisition array, a loop electrostatic coupling impedance AI adaptive identification unit, a self-diagnosis shielding integrated closed-loop control unit, a graded dynamic shielding execution module, and a local cloud-linked fault analysis and storage module;

[0010] The grounding reference module provides an equipotential reference ground for the four electrical circuits and fixes the global coupling impedance reference threshold library; the layered multi-circuit electrostatic acquisition array is attached to the cable shielding layer and PCB power pin layout of each circuit to collect four-dimensional coupled electrical parameters in real time, including circuit-to-ground electrostatic voltage, inter-circuit coupling capacitance, circuit equivalent inductive reactance, and ground leakage current.

[0011] The loop electrostatic coupling impedance AI adaptive identification unit has a built-in lightweight temporal convolution identification model, performs noise reduction and decoupling operations on the four-dimensional electrical parameters of the acquisition array, solves the dynamic values ​​of single-loop equivalent electrostatic coupling impedance and cross-loop mutual coupling impedance in real time, and completes impedance anomaly classification and labeling by benchmarking the benchmark threshold library.

[0012] The self-testing shielding integrated closed-loop control unit is the core control unit of this system. It is equipped with a self-testing shielding integrated closed-loop control program based on adaptive identification of loop electrostatic coupling impedance. It synchronously and in parallel performs online self-testing of ESD faults, dynamic tuning of shielding parameters, and closed-loop feedback compensation operations, eliminating conflicts between self-testing timing and shielding operation timing.

[0013] The graded dynamic shielding execution module is configured with four independent electrical circuits, including adjustable magnetic damping ESD shielding plates and variable impedance electrostatic discharge branches. It receives instructions from the closed-loop control unit to adjust the shielding dielectric constant, discharge impedance, and plate-to-ground bias voltage in a differentiated manner.

[0014] The local cloud-linked fault analysis and storage module binds impedance variation data, shields execution parameters, and loop operation data to complete fault source tracing and hierarchical judgment, synchronously stores closed-loop control iterative datasets, and reversely optimizes the recognition accuracy of the convolutional recognition model.

[0015] The self-testing shielded integrated closed-loop control technology based on adaptive identification of loop electrostatic coupling impedance executes as follows:

[0016] S01. Synchronous acquisition of multi-loop coupled electrical parameters: The hierarchical multi-loop electrostatic acquisition array uses a high-frequency differential sampling frequency of 1.2MHz to collect the steady-state operation coupling parameters and transient ESD impact coupling parameters of the four electrical circuits of the AI ​​robot in a time-division and interference-free manner. It eliminates the working condition noise interference caused by the start and stop of the robot servo motor and the laser sensor pulse, and outputs a decoupled and pure coupled electrical data set.

[0017] S02, Global Electrostatic Coupling Impedance Adaptive AI Identification: The temporal convolution identification model imports a pure coupled electrical dataset, splits the impedance components into two types: self-coupling impedance and mutual coupling impedance, and combines the robot's real-time posture, load power, and environmental electric field temperature and humidity factors to iteratively correct the impedance identification coefficients and accurately output the real-time global coupling impedance matrix to distinguish three types of impedance characteristics: normal operating condition impedance, electrostatic accumulation and abnormal impedance, and hardware fault distortion impedance.

[0018] S03, Impedance Linkage Integrated Parallel Self-Test: Based on the identified impedance matrix, it benchmarks against the preset four-level impedance fault thresholds and performs four types of fault self-tests in parallel: Level 1 shield plate dielectric aging fault, Level 2 circuit cable insulation damage and mutual coupling over-standard fault, Level 3 grounding module virtual connection reference drift fault, and Level 4 instantaneous ESD impact overload fault, and simultaneously generates fault location, fault level, and remaining protection life self-test message.

[0019] S04, Impedance Matching Dynamic Shielding Closed-Loop Tuning: The closed-loop control unit uses the real-time global coupling impedance matrix as the control target value, and reversely matches the resistance value of the variable impedance electrostatic discharge branch of each circuit and the bias voltage of the shielding plate. It improves the low-frequency electrostatic shielding depth for high mutual coupling communication sensing circuits and improves the transient ESD discharge rate for high-power power circuits, realizing real-time conjugate matching between shielding impedance and circuit coupling impedance.

[0020] S05, Operating Condition Feedback Closed-Loop Correction: Collect secondary circuit coupling impedance data after shielding operation, send it back to the AI ​​identification unit for difference comparison, dynamically fine-tune the shielding execution parameters, and form a complete closed loop of impedance identification - fault self-test shielding control - secondary impedance verification, so as to realize the integrated control of self-test operation and shielding operation timing and parameter linkage.

[0021] As a preferred embodiment of the present invention, the hierarchical multi-loop electrostatic acquisition array includes an isolated differential sampling chip, a high-frequency magnetically isolated coupling sensor, a voltage divider electrostatic acquisition terminal, and a time-division multiplexing frequency division circuit; each electrical loop is independently configured with a set of high-frequency magnetically isolated coupling sensors, and the four sensors are physically isolated to eliminate electrostatic crosstalk between sampling loops. The frequency division circuit divides the self-test sampling time slot and the shielded feedback sampling time slot, with a time slot interval ≤0.4μs, to achieve sampling without interference.

[0022] As a preferred embodiment of the present invention, the circuit electrostatic coupling impedance AI adaptive identification unit has a built-in working condition coupling factor database. The database contains impedance correction coefficients for four types of environmental working conditions: workshop alternating electric field, high and low temperature (-20℃~65℃), dust insulation attenuation, and robot joint torsional deformation. The temporal convolution identification model adopts a dual-channel feature extraction architecture. Channel 1 extracts steady-state low-frequency coupling impedance features, and channel 2 extracts nanosecond-level ESD transient impact impedance features. After dual-channel fusion, the identification error is ≤1.8%.

[0023] As a preferred embodiment of the present invention, the self-test and shielding integrated closed-loop control unit incorporates a time-series linkage scheduling algorithm, reuses the same set of impedance identification computing resources, eliminates the traditional independent computing power architecture of the self-test module and shielding module, and improves the computing power reuse rate by 62%. When the self-test determines that the circuit has an instantaneous ESD overload fault, the control unit switches the shielding discharge branch to the low impedance discharge level in milliseconds to complete the rapid discharge of static electricity. When the self-test determines that the shielding hardware has an aging fault, it adaptively increases the shielding compensation amplitude of the adjacent circuit to temporarily supplement the protection capability of the aging circuit and realize fault tolerance protection.

[0024] As a preferred embodiment of the present invention, the graded dynamic shielding execution module includes a numerically controlled adjustable magnetic damping shielding plate, a multi-channel programmable variable resistor discharge array, a transient TVS collaborative discharge branch, and an insulation impedance monitoring submodule; the magnetic damping shielding plate has an adjustable equivalent shielding dielectric constant range of 2.3~8.7, which is suitable for different shielding scenarios such as weak electrostatic discharge and strong ESD impact; the variable resistor discharge array has an adjustment range of 15Ω~1.2MΩ, which can accurately match the coupling conjugate impedance of different circuits.

[0025] As a preferred embodiment of the present invention, the local cloud-linked fault analysis and storage module has built-in fault classification alarm logic. Level 1 aging faults are detected locally by sound and light and recorded in the cloud ledger. Level 2 and 3 hardware faults are forcibly locked and the corresponding circuits are operated with reduced power. Level 4 ESD impact faults are instantly discharged across the entire domain and the impact impedance waveform data is retained. The cloud relies on historical impedance datasets to iteratively optimize the AI ​​identification model and adapt the impedance identification parameters of AI robots of different brands and load specifications.

[0026] As a preferred embodiment of the present invention, the four-level impedance fault threshold in S03 is calibrated based on the robot's factory no-load impedance and full-load working condition impedance as dual benchmarks, and dynamically follows the robot's service life to adaptively drift and correct, thereby eliminating the problem of fixed threshold self-test misjudgment caused by robot cable aging.

[0027] Compared with the prior art, the beneficial effects of the present invention are:

[0028] This AI robot multi-loop ESD dynamic shielding protection system, equipped with fault self-diagnosis, relies on dynamic shielding with coupling impedance conjugate matching. It improves the multi-loop cross-circuit electrostatic mutual coupling suppression rate by 51%, and the nanosecond-level ESD electrostatic discharge speed by 63%. Under complex industrial conditions of high and low temperature dust, it reduces the robot communication packet loss rate and sensor failure probability by 98%. It is suitable for all types of AI robots, including heavy-duty, high-speed, and outdoor robots.

[0029] Based on impedance characteristic AI identification self-test, it breaks through the limitations of traditional continuity self-test. It can detect cable micro-cracks, grounding impedance drift, and latent faults such as plate dielectric attenuation. The accuracy of full-domain fault self-test is ≥99.2%. It adaptively corrects the operating condition threshold and the false alarm rate of self-test is less than 0.8%. It can accurately trace the source of four types of fault points.

[0030] The integrated computing power reuse design eliminates the need for two separate hardware sets: a self-testing main controller and a shielded main controller. The overall protection module size is reduced by 42%, making it suitable for the narrow wiring space of the robot. The overall power consumption is reduced by 38%, making it suitable for the robot's battery-powered mobile operation. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the structure of a multi-loop ESD dynamic shielding protection system for an AI robot equipped with fault self-diagnosis according to the present invention.

[0032] Figure 2 This is a schematic diagram of the self-testing and shielding integrated closed-loop control execution process of an AI robot multi-loop ESD dynamic shielding protection system equipped with fault self-testing according to the present invention. Detailed Implementation

[0033] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0034] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front end," "rear end," "both ends," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0035] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0036] Please see Figure 1-2 One embodiment of the present invention is an AI robot multi-loop ESD dynamic shielding protection system equipped with fault self-diagnosis, which includes a grounding reference module, a layered multi-loop electrostatic acquisition array, an AI adaptive identification unit for loop electrostatic coupling impedance, a self-diagnosis shielding integrated closed-loop control unit, a graded dynamic shielding execution module, and a local cloud-linked fault analysis and storage module.

[0037] The grounding reference module provides equipotential grounding for four electrical circuits, unifying the potential difference across the entire domain. It has a built-in solidified global coupling impedance reference threshold library, providing reference parameters for subsequent impedance comparison and fault diagnosis. The layered multi-circuit electrostatic acquisition array is fixedly deployed on the cable shielding layer and PCB power pins of each circuit, and synchronously acquires four-dimensional coupled electrical parameters in real time across the entire domain, including circuit-to-ground electrostatic voltage, inter-circuit coupling capacitance, circuit equivalent inductive reactance, and grounding leakage current, achieving full-dimensional acquisition of electrostatic parameters.

[0038] The loop electrostatic coupling impedance AI adaptive identification unit has a built-in lightweight time-series convolution identification model. It connects to the acquisition array to transmit four-dimensional electrical parameters to complete noise reduction and decoupling calculations. It solves the dynamic values ​​of single-loop equivalent electrostatic coupling impedance and cross-loop mutual coupling impedance in real time. It benchmarks the impedance reference threshold library and automatically completes the classification and labeling of loop impedance variation levels. The self-test shielding integrated closed-loop control unit is the core control unit of the system. It is equipped with a loop electrostatic coupling impedance adaptive identification closed-loop control program. It synchronously and in parallel performs three operations: ESD fault online self-test, shielding parameter dynamic adjustment, and closed-loop feedback compensation. It completely eliminates the conflict between the self-test timing and the shielding operation timing, ensuring uninterrupted protection operation.

[0039] The hierarchical dynamic shielding execution module is equipped with adjustable magnetic damping ESD shielding plates and variable impedance electrostatic discharge branches for the robot's four electrical circuits. It independently receives control commands from the closed-loop control unit and adjusts the circuit shielding dielectric constant, electrostatic discharge impedance, and plate-to-ground bias voltage in a differentiated manner to achieve hierarchical shielding and discharge of strong and weak current circuits. The local cloud-linked fault analysis and storage module binds circuit impedance variation data, shielding execution operation parameters, and real-time circuit operation data of the robot to complete fault source tracing and hierarchical judgment. It retains the operation and maintenance log locally and uploads the closed-loop control iterative dataset to the cloud. Iterative optimization of the temporal convolutional identification model is performed in reverse to effectively reduce impedance identification errors in the long term.

[0040] In this embodiment, the self-testing shielding integrated closed-loop control technology based on adaptive identification of loop electrostatic coupling impedance specifically executes five closed-loop steps: S01, Synchronous acquisition of multi-loop coupled electrical parameters: A layered multi-loop electrostatic acquisition array uses a high-frequency differential sampling frequency of 1.2MHz to collect the steady-state operation coupling parameters and transient ESD impact coupling parameters of the four electrical loops of the AI ​​robot in a time-division, interference-free manner. The algorithm eliminates the operating condition noise interference caused by the start-stop of the robot's servo motor and laser sensor pulses, and outputs a decoupled and pure coupled electrical dataset; S02, Adaptive AI identification of global electrostatic coupling impedance: The time-series convolutional identification model imports the pure coupled electrical dataset, splits the impedance components into two types: self-coupling impedance and mutual coupling impedance, and combines the robot's real-time posture, load power, and environmental electric field temperature and humidity operating condition factors to iteratively correct the impedance identification coefficient, accurately outputting a real-time global coupled impedance matrix, and distinguishing three types of impedance characteristics: normal operating condition impedance, electrostatic accumulation abnormal impedance, and hardware fault distortion impedance; S03, Impedance linkage integrated parallel self-testing: Based on the identified impedance matrix, the system performs a self-testing process that... The system presets four levels of impedance fault thresholds and performs self-checks and judgments for four types of faults in parallel: first-level shielding plate dielectric aging fault, second-level circuit cable insulation damage and excessive mutual coupling fault, third-level grounding module loose connection and reference drift fault, and fourth-level instantaneous ESD overload fault. It simultaneously generates self-check messages for fault location, fault level, and remaining protection life. S04, Impedance Matching Dynamic Shielding Closed-Loop Setting: The closed-loop control unit uses the real-time global coupling impedance matrix as the control target value and reverse-matches the resistance of the variable impedance electrostatic discharge branch of each circuit and the shielding plate offset. The voltage setting enhances the low-frequency electrostatic shielding depth for high-coupling communication sensing circuits and improves the transient ESD discharge rate for high-power power circuits, achieving real-time conjugate matching between shielding impedance and circuit coupling impedance; S05, Operating condition feedback closed-loop correction: After shielding operation, secondary circuit coupling impedance data is collected and sent back to the AI ​​identification unit for difference comparison, dynamically fine-tuning the shielding execution parameters, forming a complete closed loop of impedance identification - fault self-test shielding control - secondary impedance verification, realizing the integration of self-test operation and shielding operation timing and parameter linkage integrated management.

[0041] In this embodiment, the hierarchical multi-loop electrostatic acquisition array includes an isolated differential sampling chip, a high-frequency magnetically isolated coupled sensor, a voltage divider electrostatic acquisition terminal, and a time-division multiplexing frequency divider circuit. Each electrical loop is independently configured with a set of high-frequency magnetically isolated coupled sensors. The four sensors are physically isolated to prevent electrostatic crosstalk between sampling loops. The frequency divider circuit divides the self-test sampling time slot and the shielded feedback sampling time slot, with a time slot interval of ≤0.4μs. The multi-loop time-division sampling does not interfere with each other, ensuring the accuracy of the sampling data.

[0042] In this embodiment, the loop electrostatic coupling impedance AI adaptive identification unit has a built-in working condition coupling factor database. The database contains impedance correction coefficients for four types of environmental working conditions: workshop alternating electric field, high and low temperature (-20℃~65℃), dust insulation attenuation, and robot joint torsional deformation. The temporal convolution identification model adopts a dual-channel feature extraction architecture. Channel 1 extracts steady-state low-frequency coupling impedance features, and channel 2 extracts nanosecond-level ESD transient impact impedance features. After dual-channel fusion, the global impedance identification error is ≤1.8%.

[0043] In this embodiment, the self-test and shielding integrated closed-loop control unit incorporates a time-series linkage scheduling algorithm, reusing the same set of impedance identification computing resources and eliminating the traditional independent computing power architecture of the self-test module and shielding module, thus improving the computing power reuse rate by 62%. When the self-test determines that the circuit has an instantaneous ESD overload fault, the control unit switches the shielding discharge branch to the low impedance discharge level in milliseconds to complete the rapid discharge of high voltage static electricity. When the self-test determines that the shielding hardware has an aging fault, it adaptively increases the shielding compensation amplitude of adjacent circuits to temporarily supplement the electrostatic protection capability of the aging circuit, achieving fault-tolerant fallback protection.

[0044] In this embodiment, the hierarchical dynamic shielding execution module includes a numerically controlled adjustable magnetic damping shielding plate, a multi-channel programmable variable resistor discharge array, a transient TVS collaborative discharge branch, and an insulation impedance monitoring submodule. The magnetic damping shielding plate has an adjustable equivalent shielding dielectric constant range of 2.3 to 8.7, which is suitable for different shielding scenarios such as weak static electricity accumulation and strong ESD high-voltage impact in workshops. The variable resistor discharge array has a resistance adjustment range of 15Ω to 1.2MΩ, which can accurately match the coupling conjugate impedance of different types of circuits to achieve optimal impedance matching discharge.

[0045] In this embodiment, the local cloud-linked fault analysis and storage module has built-in hierarchical linkage fault alarm and handling logic. For level one shielding dielectric aging faults, local audible and visual warnings are issued and cloud ledgers are archived. For level two and three cable insulation and grounding connection hardware faults, the corresponding fault circuit is forcibly locked and the power is reduced and the current is limited to avoid hardware breakdown. For level four instantaneous ESD impact faults, the entire domain parallel discharge is immediately started and the original impact impedance waveform traceability data is retained. The cloud relies on a massive historical impedance dataset to continuously iterate and optimize the AI ​​time-series convolutional identification model to adapt the impedance identification adaptation parameters of AI robots of different brands and load specifications.

[0046] In this embodiment, the four-level impedance fault threshold in the closed-loop self-test step S03 is jointly calibrated based on the robot's factory no-load impedance and full-load working condition impedance as dual benchmarks. The threshold can dynamically follow the robot's service time to adaptively drift and correct, offsetting the parameter offset caused by robot cable aging and shielding layer attenuation, thus eliminating the problem of misjudgment in fixed threshold self-test from the source.

[0047] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis, characterized in that, It includes a grounding reference module, a hierarchical multi-loop electrostatic acquisition array, an AI adaptive identification unit for loop electrostatic coupling impedance, a self-testing shielding integrated closed-loop control unit, a hierarchical dynamic shielding execution module, and a local cloud-linked fault analysis and storage module. The grounding reference module provides an equipotential reference ground for the four electrical circuits and fixes the global coupling impedance reference threshold library; the layered multi-circuit electrostatic acquisition array is attached to the cable shielding layer and PCB power pin layout of each circuit to collect four-dimensional coupled electrical parameters in real time, including circuit-to-ground electrostatic voltage, inter-circuit coupling capacitance, circuit equivalent inductive reactance, and ground leakage current. The loop electrostatic coupling impedance AI adaptive identification unit has a built-in lightweight temporal convolution identification model, performs noise reduction and decoupling operations on the four-dimensional electrical parameters of the acquisition array, solves the dynamic values ​​of single-loop equivalent electrostatic coupling impedance and cross-loop mutual coupling impedance in real time, and completes impedance anomaly classification and labeling by benchmarking the benchmark threshold library. The self-testing shielding integrated closed-loop control unit is the core control unit of this system. It is equipped with a self-testing shielding integrated closed-loop control program based on adaptive identification of loop electrostatic coupling impedance. It synchronously and in parallel performs online self-testing of ESD faults, dynamic tuning of shielding parameters, and closed-loop feedback compensation operations, eliminating conflicts between self-testing timing and shielding operation timing. The graded dynamic shielding execution module is configured with four independent electrical circuits, including adjustable magnetic damping ESD shielding plates and variable impedance electrostatic discharge branches. It receives instructions from the closed-loop control unit to adjust the shielding dielectric constant, discharge impedance, and plate-to-ground bias voltage in a differentiated manner. The local cloud-linked fault analysis and storage module binds impedance variation data, shields execution parameters, and loop operation data to complete fault source tracing and hierarchical judgment, synchronously stores closed-loop control iterative datasets, and reversely optimizes the recognition accuracy of the convolutional recognition model.

2. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The self-testing shielded integrated closed-loop control technology based on adaptive identification of loop electrostatic coupling impedance executes as follows: S01. Synchronous acquisition of multi-loop coupled electrical parameters: The hierarchical multi-loop electrostatic acquisition array uses a high-frequency differential sampling frequency of 1.2MHz to collect the steady-state operation coupling parameters and transient ESD impact coupling parameters of the four electrical circuits of the AI ​​robot in a time-division and interference-free manner. It eliminates the working condition noise interference caused by the start and stop of the robot servo motor and the laser sensor pulse, and outputs a decoupled and pure coupled electrical data set. S02, Global Electrostatic Coupling Impedance Adaptive AI Identification: The temporal convolution identification model imports a pure coupled electrical dataset, splits the impedance components into two types: self-coupling impedance and mutual coupling impedance, and combines the robot's real-time posture, load power, and environmental electric field temperature and humidity factors to iteratively correct the impedance identification coefficients and accurately output the real-time global coupling impedance matrix to distinguish three types of impedance characteristics: normal operating condition impedance, electrostatic accumulation and abnormal impedance, and hardware fault distortion impedance. S03, Impedance Linkage Integrated Parallel Self-Test: Based on the identified impedance matrix, it benchmarks against the preset four-level impedance fault thresholds and performs four types of fault self-tests in parallel: Level 1 shield plate dielectric aging fault, Level 2 circuit cable insulation damage and mutual coupling over-standard fault, Level 3 grounding module virtual connection reference drift fault, and Level 4 instantaneous ESD impact overload fault, and simultaneously generates fault location, fault level, and remaining protection life self-test message. S04, Impedance Matching Dynamic Shielding Closed-Loop Tuning: The closed-loop control unit uses the real-time global coupling impedance matrix as the control target value, and reversely matches the resistance value of the variable impedance electrostatic discharge branch of each circuit and the bias voltage of the shielding plate. It improves the low-frequency electrostatic shielding depth for high mutual coupling communication sensing circuits and improves the transient ESD discharge rate for high-power power circuits, realizing real-time conjugate matching between shielding impedance and circuit coupling impedance. S05, Operating Condition Feedback Closed-Loop Correction: Collect secondary circuit coupling impedance data after shielding operation, send it back to the AI ​​identification unit for difference comparison, dynamically fine-tune the shielding execution parameters, and form a complete closed loop of impedance identification - fault self-test shielding control - secondary impedance verification, so as to realize the integrated control of self-test operation and shielding operation timing and parameter linkage.

3. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The hierarchical multi-loop electrostatic acquisition array includes an isolated differential sampling chip, a high-frequency magnetically isolated coupling sensor, a voltage divider electrostatic acquisition terminal, and a time-division multiplexing frequency divider circuit. Each electrical loop is independently configured with a set of high-frequency magnetically isolated coupling sensors. The four sensors are physically isolated to eliminate electrostatic crosstalk between sampling loops. The frequency divider circuit divides the self-test sampling time slot and the shielded feedback sampling time slot, with a time slot interval of ≤0.4μs, to achieve sampling without interference.

4. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The circuit electrostatic coupling impedance AI adaptive identification unit has a built-in working condition coupling factor database. The database contains impedance correction coefficients for four types of environmental working conditions: workshop alternating electric field, high and low temperature (-20℃~65℃), dust insulation attenuation, and robot joint torsional deformation. The temporal convolution identification model adopts a dual-channel feature extraction architecture. Channel 1 extracts steady-state low-frequency coupling impedance features, and channel 2 extracts nanosecond-level ESD transient impact impedance features. After dual-channel fusion, the identification error is ≤1.8%.

5. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The integrated self-test and shielding closed-loop control unit incorporates a time-series linkage scheduling algorithm, reusing the same set of impedance identification computing resources and eliminating the traditional independent computing architecture of the self-test module and shielding module, thus improving the computing power reuse rate by 62%. When the self-test determines that the circuit has an instantaneous ESD overload fault, the control unit switches the shielding discharge branch to the low impedance discharge level in milliseconds to complete the rapid discharge of static electricity. When the self-test determines that the shielding hardware has an aging fault, it adaptively increases the shielding compensation amplitude of adjacent circuits to temporarily supplement the protection capability of the aging circuit and achieve fault tolerance protection.

6. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The hierarchical dynamic shielding execution module includes a numerically controlled adjustable magnetic damping shielding plate, a multi-channel programmable variable resistor discharge array, a transient TVS collaborative discharge branch, and an insulation impedance monitoring submodule. The magnetic damping shielding plate has an adjustable equivalent shielding dielectric constant range of 2.3 to 8.7, which is suitable for different shielding scenarios such as weak electrostatic discharge and strong ESD impact. The variable resistor discharge array has an adjustment range of 15Ω to 1.2MΩ, which can accurately match the coupling conjugate impedance of different circuits.

7. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 1, characterized in that, The local cloud-linked fault analysis and storage module has built-in fault classification alarm logic. Level 1 aging faults are detected locally by sound and light and recorded in the cloud ledger. Level 2 and 3 hardware faults are forcibly locked and the corresponding circuit is operated with reduced power. Level 4 ESD impact faults are discharged in real time and the impact impedance waveform data is retained. The cloud relies on historical impedance datasets to iteratively optimize the AI ​​identification model and adapt the impedance identification parameters of AI robots of different brands and load specifications.

8. The multi-loop ESD dynamic shielding protection system for AI robots equipped with fault self-diagnosis as described in claim 2, characterized in that, The four-level impedance fault threshold in S03 is based on dual benchmark calibration of the robot's factory no-load impedance and full-load working condition impedance, and dynamically follows the robot's service time to adaptively drift and correct, eliminating the problem of fixed threshold self-test misjudgment caused by robot cable aging.

Images