Building intercom system full-scene automatic test platform and method based on AI drive

Abstract

The invention discloses a building intercom system full-scene automatic test platform and method based on AI driving. The test platform comprises an intelligent control host, an AI algorithm engine, a big data analysis platform, an Internet of Things test terminal, an environment simulation cabin, an excitation generation module, a communication module, a display and operation module and a storage module. The intelligent control host comprises an ARM chip, a real-time operating system and a control module. Through cooperation of the Internet of Things test terminal and the AI algorithm engine, the purposes of automatic cycle linkage test and full-link closed-loop test can be achieved, the problem of single function module test in the prior art is solved, the test path can be automatically generated according to the community house type diagram, and the test efficiency is improved. The effect of simulating a real cell topological structure and a user behavior track for testing is achieved, and the use requirement is met.

Description

Technical Field

[0001] The present invention relates to the technical field of full-scenario automated testing of building intercom systems, and in particular to an AI-driven full-scenario automated testing platform and method for building intercom systems. Background Art

[0002] After a period of use, the building intercom system needs to be tested regularly to ensure its stability during use. Traditional building intercom system testing relies on manual simulation of calling, unlocking, alarming and other scenarios, which is inefficient and prone to missed detections. In addition, multi-device linkage testing requires manual point-by-point debugging, and the compatibility verification cost is high. To solve this problem, an automated testing device has appeared on the market, but the automated testing device still has the following deficiencies during use: 1. Most automated testing equipment is a single functional module and lacks full-link closed-loop testing capabilities; 2. It is unable to simulate the real cell topology and user behavior trajectory, and the test scenario is single; Based on the above, this application proposes an AI-driven full-scenario automated testing platform and method for building intercom systems. Summary of the Invention

[0003] Based on the technical problems existing in the background technology, the present invention proposes an AI-driven full-scene automated testing platform and method for building intercom systems.

[0004] The present invention proposes an AI-driven full-scenario automated testing platform for building intercom systems, which includes an intelligent control host, an AI algorithm engine, a big data analysis platform, an Internet of Things test terminal, an environmental simulation cabin, an incentive generation module, a communication module, a display and operation module, and a storage module.

[0005] The intelligent control host includes an ARM chip, a real-time operating system (RTOS) and a control module;

[0006] The environmental simulation cabin includes a data acquisition module;

[0007] The control module is connected to the data acquisition module, the stimulus generation module, and the storage module through a high-speed data bus; the communication module is connected to the control module through a corresponding interface protocol and establishes a communication link with the device under test and external devices; the display and operation module is connected to the control module through an HDMI, VGA video interface, and a USB control interface.

[0008] Preferably, the control module is composed of a high-performance industrial-grade controller or server with powerful data processing and logical operation capabilities. It is responsible for the operation scheduling of the entire test platform. According to the preset test process and strategy, it sends instructions to the stimulus generation module to generate stimulus signals, controls the acquisition frequency and acquisition duration parameters of the data acquisition module, analyzes and processes the collected data, and stores the test results in the storage module. At the same time, it controls the display and operation module to display test information.

[0009] Preferably, the IoT test terminal supports Zigbee / 4G / 5G / Wi-Fi multi-protocol communications and simulates resident terminals, management centers, and third-party devices;

[0010] The environmental simulation chamber integrates temperature and humidity / light / electromagnetic interference generating devices, supports 20°C-80°C temperature, 0-95% humidity and 0-100V / m electromagnetic radiation adjustment. The data acquisition module includes various sensors and signal acquisition card devices, which can collect multiple physical quantity data such as voltage, current, temperature and pressure during the test in real time, as well as digital signals and communication protocol data. The data acquisition module is used to convert various physical quantity signals and digital signals generated by the device under test under the action of the excitation signal into computer-recognizable digital signals, and transmit them to the control module to provide raw data for test analysis.

[0011] Preferably, the AI ​​algorithm engine is based on the TensorFlow Lite framework, with a built-in fault prediction model and scenario generation algorithm. The fault prediction model is trained with historical test data and has an accuracy rate of ≥92%. The scenario generation algorithm automatically generates a test path based on the residential area layout.

[0012] The big data analysis platform is used to store test logs and generate visual reports, including pass rate trends, high-frequency failure distribution, and performance bottleneck heat maps.

[0013] Preferably, the excitation generation module can generate different types of excitation signals, including an analog signal generator that can output sine wave and square wave analog signals, and a digital signal generator that can generate a specific digital pulse sequence. The excitation generation module can generate an excitation signal that meets the test requirements in accordance with the instructions of the control module, apply it to the device under test, and simulate various input conditions of the device in the actual working environment.

[0014] Preferably, the communication module includes an Ethernet interface, a serial port, a USB interface, and a wireless communication module, and supports data interaction with the device under test, external devices, and a remote server. The communication module is used to realize communication between the test platform and the device under test, so as to control the working status of the device under test and obtain the internal parameters of the device. At the same time, it supports data interaction between the test platform and external devices and remote servers, and realizes remote transmission, remote monitoring, and remote management of test data.

[0015] The display and operation module is a combination of a touch screen, a display, a keyboard and a mouse, which facilitates the operator to set test parameters, control the test process, and view test data, waveforms and test result reports in real time; the display and operation module is used to receive display data transmitted by the control module and feed back the operator's instructions to the control module, providing the operator with a friendly human-computer interaction interface, facilitating the operator to set test parameters, start, pause and stop the test process, observe the change trend and waveform of the test data in real time, view the test result report, and perform printing and export operations. Its test parameters include the type, amplitude, frequency of the excitation signal and the trigger conditions for data acquisition.

[0016] Preferably, the storage module is a large-capacity hard disk, a solid-state drive (SSD) or a cloud storage service, which is used to store the original data, intermediate processed data and final test report generated during the test process, and to completely save the original data during the test process for subsequent data tracing and in-depth analysis. At the same time, it can store processed test result data and test reports to provide data support for product quality assessment and fault diagnosis.

[0017] The present invention also proposes an AI-driven full-scenario automated testing method for a building intercom system, comprising the following steps:

[0018] S1: The operator opens the test platform software through the display and operation module and enters the test project configuration interface. Based on the test requirements of the device under test, the operator sets the test process and test parameters in the control module. At the same time, the operator imports the CAD community floor plan and automatically generates the logical topology of buildings, units, and residents to simulate the real networking environment. The AI ​​algorithm engine automatically generates the test path based on the community floor plan based on the scenario generation algorithm;

[0019] S2: Establish a connection between the test platform and the device under test through the communication module to ensure normal communication. If the device under test supports Ethernet communication, use a network cable to connect the Ethernet interface of the test platform to the network port of the device under test, and configure the corresponding IP address and port number communication parameters;

[0020] S3: The control module sends instructions to the stimulus generation module based on the set test parameters. The stimulus generation module generates the corresponding stimulus signal and applies it to the device under test through the connecting line to simulate various input conditions of the device in the actual working environment.

[0021] S4: After the excitation signal is applied to the device under test, the device under test generates a response signal. The data acquisition module collects the response signal of the device under test in real time according to the acquisition parameters set by the control module, converts the collected analog signal into a digital signal, and transmits it to the control module through the data bus;

[0022] S5: After receiving the data transmitted by the data acquisition module, the control module performs real-time analysis and processing on the data, filters the collected voltage signal to remove noise interference, calculates the amplitude, frequency and phase characteristic parameters of the signal, and compares them with the preset standard values ​​to determine whether the performance of the device under test meets the requirements;

[0023] S6: The control module transmits the real-time processed data to the display and operation module, which is displayed to the operator in the form of numbers and waveforms on the interface, making it convenient for the operator to monitor the test process in real time. At the same time, the control module stores the original data and processed data in the storage module;

[0024] S7: At the same time, the AI ​​algorithm engine performs automatic cycle testing, linkage testing, and stress testing, and determines whether the device under test has a fault based on vibration sensor and current waveform analysis. If so, a fault location report is generated. If not, a pass report is generated and transmitted to the control module. During the test process, the fault prediction model automatically increases the test frequency of high-risk scenarios based on historical fault data, and analyzes the test data through algorithms to provide early warning of equipment aging trends and assist in formulating maintenance plans.

[0025] S8: When the preset test end condition is reached, the control module sends a stop instruction to the stimulus generation module to stop the output of the stimulus signal, and at the same time controls the data acquisition module to stop data acquisition;

[0026] S9: The control module comprehensively analyzes and processes the test data in the storage module and generates a detailed test result report. The report includes test items, test parameters, test data, performance index evaluation results, and a conclusion on whether the test is qualified.

[0027] S10: The control module transmits the test result report and the report generated by the AI ​​algorithm engine to the display and operation module. The operator can view the report content on the interface and print and export it as needed. The test result report is automatically sent to the remote server or the mailbox of the relevant manager through the communication module to realize the sharing and remote management of the test results.

[0028] Preferably, in S7, the fault prediction model in the AI ​​algorithm engine needs to be trained and verified before use, and the specific logical steps are as follows:

[0029] S701: Perform data cleaning and feature engineering operations on the data collected by the data acquisition module, delete outliers, and extract features of vibration frequency, current waveform, and corresponding delay;

[0030] S702: Divide the collected data set into a 70% training set, a 20% validation set, and a 10% test set;

[0031] S703: Select the fault prediction model of the LSTM+Attention mechanism as the training model;

[0032] S704: Using the training set and the loss function to train the selected model, and using the validation set to train the model;

[0033] S705: Label the training indicators. If they meet the requirements, directly deploy the model to the algorithm engine. Otherwise, use the regularization coefficient to adjust the hyperparameters and continue to iterate the training and verification operations.

[0034] Compared with the existing technology, the beneficial effects of the present invention are:

[0035] The present invention cooperates with the Internet of Things test terminal and the AI ​​algorithm engine to realize automatic cyclic linkage testing and achieve the purpose of full-link closed-loop testing, solving the problem of single functional module testing in the existing technology, and can automatically generate a test path according to the community floor plan, achieving the effect of simulating the real community topology structure and user behavior trajectory for testing, thus meeting usage requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a system block diagram of the AI-driven full-scenario automated testing platform for building intercom systems proposed in this invention;

[0037] Figure 2 This is a flowchart of the AI-driven full-scenario automated testing method for building intercom systems proposed in this invention;

[0038] Figure 3 This is a flowchart of the fault prediction model training in the AI-driven full-scenario automated testing method for building intercom systems proposed in the present invention. DETAILED DESCRIPTION

[0039] The present invention will be further explained below with reference to specific embodiments.

[0040] Example

[0041] Reference Figure 1-3 ,This embodiment proposes an AI-driven full-scenario automated testing platform for building intercom systems,,including an intelligent control host, an AI algorithm engine, a big data analysis platform, an IoT test terminal, an environmental simulation cabin, a stimulus generation module, a communication module, a display and operation module, and a storage module;

[0042] IoT test terminal - supports Zigbee / 4G / 5G / Wi-Fi multi-protocol communications, simulating resident terminals, management centers, and third-party devices;

[0043] The AI ​​algorithm engine is based on the TensorFlow Lite framework and has a built-in fault prediction model and scenario generation algorithm. The fault prediction model is trained using historical test data and has an accuracy rate of ≥92%. The scenario generation algorithm automatically generates test paths based on the residential floor plan.

[0044] The big data analysis platform is used to store test logs and generate visual reports, including pass rate trends, high-frequency failure distribution, and performance bottleneck heat maps;

[0045] The intelligent control host includes an ARM chip, a real-time operating system (RTOS), and a control module;

[0046] The control module is composed of a high-performance industrial-grade controller or server with powerful data processing and logical operation capabilities. It is responsible for the operation and scheduling of the entire test platform. According to the preset test process and strategy, it sends instructions to the stimulus generation module to generate stimulus signals, controls the acquisition frequency and acquisition duration parameters of the data acquisition module, analyzes and processes the collected data, and stores the test results in the storage module. At the same time, it controls the display and operation module to display test information.

[0047] The environmental simulation chamber includes a data acquisition module; the environmental simulation chamber integrates temperature and humidity / light / electromagnetic interference generation devices, supporting 20°C-80°C temperature, 0-95% humidity, and 0-100V / m electromagnetic radiation adjustment. The data acquisition module contains various sensors and signal acquisition card devices, which can collect multiple physical quantity data such as voltage, current, temperature, and pressure during the test in real time, as well as digital signals and communication protocol data. The data acquisition module is used to convert the various physical quantity signals and digital signals generated by the device under test under the action of the excitation signal into computer-recognizable digital signals, and transmit them to the control module to provide raw data for test analysis;

[0048] The control module is connected to the data acquisition module, stimulus generation module, and storage module through a high-speed data bus. The communication module is connected to the control module through the corresponding interface protocol and establishes a communication link with the device under test and external devices. The display and operation module is connected to the control module through the HDMI, VGA video interface and USB control interface.

[0049] The stimulus generation module can generate different types of stimulus signals, including analog signal generators that can output sine and square wave analog signals, and digital signal generators that can generate specific digital pulse sequences. The stimulus generation module can generate stimulus signals that meet the test requirements according to the instructions of the control module and apply them to the device under test to simulate various input conditions of the device in the actual working environment.

[0050] The communication module includes Ethernet interface, serial port, USB interface, and wireless communication module, supporting data interaction with the device under test, external devices, and remote servers. The communication module is used to realize communication between the test platform and the device under test, so as to control the working status of the device under test and obtain the internal parameters of the device. At the same time, it supports data interaction between the test platform and external devices and remote servers, realizing remote transmission, remote monitoring, and remote management of test data.

[0051] The display and operation module is a combination of a touch screen, a display, a keyboard, and a mouse, which allows operators to set test parameters, control the test process, and view test data, waveforms, and test result reports in real time. The display and operation module is used to receive display data transmitted by the control module and feed back the operator's instructions to the control module, providing the operator with a friendly human-computer interaction interface, allowing the operator to set test parameters, start, pause, and stop the test process, observe the changing trends and waveforms of test data in real time, view test result reports, and print and export them. Its test parameters include the type, amplitude, and frequency of the excitation signal, as well as the trigger conditions for data acquisition.

[0052] The storage module is a large-capacity hard disk, solid-state drive (SSD) or cloud storage service. It is used to store the original data, intermediate processing data and final test report generated during the test process. It also completely saves the original data during the test process to facilitate subsequent data tracing and in-depth analysis. At the same time, it can store processed test result data and test reports to provide data support for product quality assessment and fault diagnosis.

[0053] This embodiment also proposes an AI-driven full-scenario automated testing method for a building intercom system, including the following steps:

[0054] S1: The operator opens the test platform software through the display and operation module and enters the test project configuration interface. Based on the test requirements of the device under test, the operator sets the test process and test parameters in the control module. At the same time, the operator imports the CAD community floor plan and automatically generates the logical topology of buildings, units, and residents to simulate the real networking environment. The AI ​​algorithm engine automatically generates the test path based on the community floor plan based on the scenario generation algorithm;

[0055] S2: Establish a connection between the test platform and the device under test through the communication module to ensure normal communication. If the device under test supports Ethernet communication, use a network cable to connect the Ethernet interface of the test platform to the network port of the device under test, and configure the corresponding IP address and port number communication parameters;

[0056] S3: The control module sends instructions to the stimulus generation module based on the set test parameters. The stimulus generation module generates the corresponding stimulus signal and applies it to the device under test through the connecting line to simulate various input conditions of the device in the actual working environment.

[0057] S4: After the excitation signal is applied to the device under test, the device under test generates a response signal. The data acquisition module collects the response signal of the device under test in real time according to the acquisition parameters set by the control module, converts the collected analog signal into a digital signal, and transmits it to the control module through the data bus;

[0058] S5: After receiving the data transmitted by the data acquisition module, the control module performs real-time analysis and processing on the data, filters the collected voltage signal to remove noise interference, calculates the amplitude, frequency and phase characteristic parameters of the signal, and compares them with the preset standard values ​​to determine whether the performance of the device under test meets the requirements;

[0059] S6: The control module transmits the real-time processed data to the display and operation module, which is displayed to the operator in the form of numbers and waveforms on the interface, making it convenient for the operator to monitor the test process in real time. At the same time, the control module stores the original data and processed data in the storage module;

[0060] S7: At the same time, the AI ​​algorithm engine performs automatic cycle testing, linkage testing, and stress testing. Based on vibration sensors and current waveform analysis, it locates hidden faults such as poor hardware contact and abnormal power ripple, and determines whether the device under test has a fault. If so, a fault location report is generated. If so, a pass report is generated and transmitted to the control module. During the test process, the fault prediction model automatically increases the test frequency of high-risk scenarios based on historical fault data, and analyzes the test data through algorithms to provide early warning of equipment aging trends and assist in formulating maintenance plans.

[0061] In addition, the automated loop test includes call connection rate, video clarity, and voice noise reduction. The linkage test process triggers the preset "resident alarm → management center pop-up window → automatic door unlocking → camera tracking" process to verify cross-system response time. The stress test simulates 1,000+ concurrent calls through IoT terminals to test system throughput and latency thresholds.

[0062] The fault prediction model in the AI ​​algorithm engine needs to be trained and verified before use. The specific logical steps are as follows:

[0063] S701: Perform data cleaning and feature engineering operations on the data collected by the data acquisition module, delete outliers, and extract features of vibration frequency, current waveform, and corresponding delay;

[0064] S702: Divide the collected data set into a 70% training set, a 20% validation set, and a 10% test set;

[0065] S703: Select the fault prediction model of the LSTM+Attention mechanism as the training model;

[0066] S704: Using the training set and the loss function to train the selected model, and using the validation set to train the model;

[0067] S705: Mark the training indicators. If they meet the requirements, the model is directly deployed to the algorithm engine. Otherwise, the regularization coefficient is used to adjust the hyperparameters, and the training and verification operations are continuously iterated.

[0068] S8: When the preset test end condition is reached, the control module sends a stop instruction to the stimulus generation module to stop the output of the stimulus signal, and at the same time controls the data acquisition module to stop data acquisition;

[0069] S9: The control module comprehensively analyzes and processes the test data in the storage module and generates a detailed test result report. The report includes test items, test parameters, test data, performance index evaluation results, and a conclusion on whether the test is qualified.

[0070] S10: The control module transmits the test result report and the report generated by the AI ​​algorithm engine to the display and operation module. The operator can view the report content on the interface and print and export it as needed. The test result report is automatically sent to the remote server or the mailbox of the relevant manager through the communication module to realize the sharing and remote management of the test results.

[0071] The above are only preferred specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. An AI-driven full-scenario automated testing platform for building intercom systems, characterized by: It includes an intelligent control host, AI algorithm engine, big data analysis platform, IoT test terminal, environmental simulation cabin, stimulus generation module, communication module, display and operation module, and storage module; The intelligent control host includes an ARM chip, a real-time operating system and a control module; The environmental simulation cabin includes a data acquisition module; The control module is connected to the data acquisition module, the stimulus generation module, and the storage module through a high-speed data bus; the communication module is connected to the control module through a corresponding interface protocol and establishes a communication link with the device under test and external devices; the display and operation module is connected to the control module through an HDMI, VGA video interface, and a USB control interface.

2. The AI-driven full-scenario automated testing platform for building intercom systems according to claim 1 is characterized in that: The control module is composed of a high-performance industrial-grade controller or server with powerful data processing and logical operation capabilities. It is responsible for the operation scheduling of the entire test platform. According to the preset test process and strategy, it sends instructions to the stimulus generation module to generate stimulus signals, controls the acquisition frequency and acquisition duration parameters of the data acquisition module, analyzes and processes the collected data, and stores the test results in the storage module. At the same time, it controls the display and operation module to display test information.

3. The AI-driven full-scenario automated testing platform for building intercom systems according to claim 1 is characterized in that: The IoT test terminal supports Zigbee / 4G / 5G / Wi-Fi multi-protocol communications and simulates resident terminals, management centers, and third-party devices; The environmental simulation chamber integrates temperature and humidity / light / electromagnetic interference generating devices, supports 20°C-80°C temperature, 0-95% humidity and 0-100V / m electromagnetic radiation adjustment. The data acquisition module includes various sensors and signal acquisition card devices, which can collect multiple physical quantity data such as voltage, current, temperature and pressure during the test in real time, as well as digital signals and communication protocol data. The data acquisition module is used to convert various physical quantity signals and digital signals generated by the device under test under the action of the excitation signal into computer-recognizable digital signals, and transmit them to the control module to provide raw data for test analysis.

4. The AI-driven building intercom system full-scenario automated testing platform according to claim 1 is characterized in that: The AI ​​algorithm engine is based on the TensorFlow Lite framework and has a built-in fault prediction model and scenario generation algorithm. The fault prediction model is trained using historical test data with an accuracy rate of ≥92%. The scenario generation algorithm automatically generates test paths based on the residential floor plan. The big data analysis platform is used to store test logs and generate visual reports, including pass rate trends, high-frequency failure distribution, and performance bottleneck heat maps.

5. The AI-driven full-scenario automated testing platform for building intercom systems according to claim 1 is characterized in that: The stimulus generation module can generate different types of stimulus signals, including an analog signal generator that can output sine wave and square wave analog signals, and a digital signal generator that can generate specific digital pulse sequences. The stimulus generation module can generate stimulus signals that meet the test requirements according to the instructions of the control module, apply them to the device under test, and simulate various input conditions of the device in the actual working environment.

6. The AI-driven building intercom system full-scenario automated testing platform according to claim 1 is characterized in that: The communication module includes an Ethernet interface, a serial port, a USB interface, and a wireless communication module, and supports data interaction with the device under test, external devices, and a remote server. The communication module is used to realize communication between the test platform and the device under test, so as to control the working status of the device under test and obtain the internal parameters of the device. At the same time, it supports data interaction between the test platform and external devices and remote servers, and realizes remote transmission, remote monitoring, and remote management of test data; The display and operation module is a combination of a touch screen, a display, a keyboard and a mouse, which facilitates the operator to set test parameters, control the test process, and view test data, waveforms and test result reports in real time; the display and operation module is used to receive display data transmitted by the control module and feed back the operator's instructions to the control module, providing the operator with a friendly human-computer interaction interface, facilitating the operator to set test parameters, start, pause and stop the test process, observe the change trend and waveform of the test data in real time, view the test result report, and perform printing and export operations. Its test parameters include the type, amplitude, frequency of the excitation signal and the trigger conditions for data acquisition.

7. The AI-driven building intercom system full-scenario automated testing platform according to claim 1 is characterized in that: The storage module is a large-capacity hard disk, solid-state drive or cloud storage service, which is used to store the original data, intermediate processed data and final test report generated during the test process, and completely save the original data during the test process to facilitate subsequent data tracing and in-depth analysis. At the same time, it can store processed test result data and test reports to provide data support for product quality assessment and fault diagnosis.

8. The AI-driven full-scenario automated testing method for building intercom systems according to any one of claims 1 to 7, characterized in that: The following steps are involved: S1: The operator opens the test platform software through the display and operation module and enters the test project configuration interface. Based on the test requirements of the device under test, the operator sets the test process and test parameters in the control module. At the same time, the operator imports the CAD community floor plan and automatically generates the logical topology of buildings, units, and residents to simulate the real networking environment. The AI ​​algorithm engine automatically generates the test path based on the community floor plan based on the scenario generation algorithm; S2: Establish a connection between the test platform and the device under test through the communication module to ensure normal communication. The device under test supports Ethernet communication. Use a network cable to connect the Ethernet interface of the test platform to the network port of the device under test, and configure the corresponding IP address and port number communication parameters. S3: The control module sends instructions to the stimulus generation module based on the set test parameters. The stimulus generation module generates the corresponding stimulus signal and applies it to the device under test through the connecting line to simulate various input conditions of the device in the actual working environment. S4: After the excitation signal is applied to the device under test, the device under test generates a response signal. The data acquisition module collects the response signal of the device under test in real time according to the acquisition parameters set by the control module, converts the collected analog signal into a digital signal, and transmits it to the control module through the data bus; S5: After receiving the data transmitted by the data acquisition module, the control module performs real-time analysis and processing on the data, filters the collected voltage signal to remove noise interference, calculates the amplitude, frequency and phase characteristic parameters of the signal, and compares them with the preset standard values ​​to determine whether the performance of the device under test meets the requirements; S6: The control module transmits the real-time processed data to the display and operation module, which is displayed to the operator in the form of numbers and waveforms on the interface, making it convenient for the operator to monitor the test process in real time. At the same time, the control module stores the original data and processed data in the storage module; S7: At the same time, the AI ​​algorithm engine performs automatic cycle testing, linkage testing, and stress testing, and determines whether the device under test has a fault based on vibration sensor and current waveform analysis. If so, a fault location report is generated. If not, a pass report is generated and transmitted to the control module. During the test process, the fault prediction model automatically increases the test frequency of high-risk scenarios based on historical fault data, and analyzes the test data through algorithms to provide early warning of equipment aging trends and assist in formulating maintenance plans. S8: When the preset test end condition is reached, the control module sends a stop instruction to the stimulus generation module to stop the output of the stimulus signal, and at the same time controls the data acquisition module to stop data acquisition; S9: The control module comprehensively analyzes and processes the test data in the storage module and generates a detailed test result report. The report includes test items, test parameters, test data, performance index evaluation results, and a conclusion on whether the test is qualified. S10: The control module transmits the test result report and the report generated by the AI ​​algorithm engine to the display and operation module. The operator can view the report content on the interface and print and export it as needed. The test result report is automatically sent to the remote server or the mailbox of the relevant manager through the communication module to realize the sharing and remote management of the test results.

9. The AI-driven full-scenario automated testing method for building intercom systems according to claim 8 is characterized in that: In S7, the fault prediction model in the AI ​​algorithm engine needs to be trained and verified before use. The specific logical steps are as follows: S701: Perform data cleaning and feature engineering operations on the data collected by the data acquisition module, delete outliers, and extract features of vibration frequency, current waveform, and corresponding delay; S702: Divide the collected data set into a 70% training set, a 20% validation set, and a 10% test set; S703: Select the fault prediction model of the LSTM+Attention mechanism as the training model; S704: Using the training set and the loss function to train the selected model, and using the validation set to train the model; S705: Label the training indicators. If they meet the requirements, directly deploy the model to the algorithm engine. Otherwise, use the regularization coefficient to adjust the hyperparameters and continue to iterate the training and verification operations.

Images