Switchgear discharge fault detection method and system

Abstract

The present application relates to the technical field of safety monitoring of high-voltage insulation equipment in power systems, and specifically discloses a switch cabinet discharge fault detection method and system, the method comprising: an analog-to-digital conversion module of a single-chip microcomputer reads the voltage of a Hosstone bridge, and sends the read first voltage value to a cloud server, wherein the Hosstone bridge is arranged in a target switch cabinet, and the Hosstone bridge includes an ozone sensor, and when a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes; after receiving the first voltage value, the cloud server inputs the first voltage value into a preset ozone concentration recognition model, obtains the current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet according to the current ozone concentration value. The present application realizes real-time intelligent detection of discharge faults in target switch cabinets, and at a low cost.

Description

Technical Field

[0001] The present application relates to the technical field of safety monitoring of high-voltage insulation equipment in power systems, and in particular to a method and system for detecting discharge faults in switch cabinets. Background Art

[0002] A large number of electrical equipment are often needed in the production of industrial equipment. For electrical equipment, insulation failure in electrical equipment is an important cause of damage to electrical equipment. These insulation failures often originate from insulation degradation caused by partial discharge. Before the through discharge occurs in the insulation of the equipment, partial discharge often occurs first. It can be said that partial discharge is both a sign of insulation failure in electrical equipment and a key factor causing insulation failure. Therefore, timely detection of partial discharge in electrical equipment is crucial to ensure the normal operation of electrical equipment.

[0003] High-voltage switchgear is a very important electrical equipment in the power system. Discharge will occur during the incubation period of its fault. In order to monitor the insulation status of high-voltage switchgear, people have developed a series of partial discharge monitoring systems for high-voltage switchgear. These monitoring systems mainly use transient ground voltage method and ultrasonic method. However, these methods cannot monitor the discharge situation in the switchgear in real time. In addition, these monitoring systems are complex and costly. Summary of the invention

[0004] In view of this, the present application provides a switch cabinet discharge fault detection method and system, which realizes real-time intelligent detection of target switch cabinet discharge faults by real-time monitoring of ozone concentration changes in the target switch cabinet and utilizing the collaborative work of a Hoston bridge, a single-chip microcomputer, a cloud server and an ozone concentration recognition model; by using low-cost devices such as a single-chip microcomputer and a Hoston bridge with an ozone sensor, and using a cloud server as a computing terminal, hardware costs are saved, space occupancy in the switch cabinet is reduced, and the complexity and cost of the monitoring system can be greatly reduced.

[0005] According to one aspect of the present application, a switch cabinet discharge fault detection method is provided, comprising:

[0006] The analog-to-digital conversion module of the single-chip microcomputer reads the voltage of the Hothstone bridge and sends the read first voltage value to the cloud server, wherein the Hothstone bridge is arranged in the target switch cabinet, and the Hothstone bridge includes an ozone sensor. When a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes;

[0007] After receiving the first voltage value, the cloud server inputs the first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet based on the current ozone concentration value.

[0008] According to another aspect of the present application, a switch cabinet discharge fault detection system is provided, comprising:

[0009] A single-chip computer, used for reading a voltage of a Hothstone bridge through an analog-to-digital conversion module, and sending the read first voltage value to a cloud server, wherein the Hothstone bridge is arranged in a target switch cabinet, and the Hothstone bridge includes an ozone sensor, and when a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes;

[0010] The cloud server is used to input the first voltage value into a preset ozone concentration identification model after receiving the first voltage value, obtain a current ozone concentration value in the target switch cabinet, and determine whether a discharge fault occurs in the target switch cabinet based on the current ozone concentration value.

[0011] By means of the above technical scheme, the present application provides a method and system for detecting discharge faults in a switch cabinet, wherein the single-chip microcomputer reads the voltage of the Hosstone bridge through its built-in analog-to-digital conversion module to obtain a first voltage value. The single-chip microcomputer then sends the first voltage value to the cloud server through the network for further processing. After receiving the first voltage value sent by the single-chip microcomputer, the cloud server can input it into a preset ozone concentration recognition model to calculate the current ozone concentration value in the target switch cabinet. Afterwards, the cloud server can determine whether a discharge fault has occurred in the target switch cabinet based on the calculated current ozone concentration value. The embodiment of the present application realizes real-time intelligent detection of discharge faults in the target switch cabinet by real-time monitoring of the change of ozone concentration in the target switch cabinet, and by using the collaborative work of the Hosstone bridge, the single-chip microcomputer, the cloud server and the ozone concentration recognition model; by using low-cost devices such as the single-chip microcomputer and the Hosstone bridge with an ozone sensor, and using the cloud server as a computing terminal, the hardware cost is saved, the space occupied in the switch cabinet is reduced, and the complexity and cost of the monitoring system can be greatly reduced.

[0012] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0014] Figure 1 A schematic diagram of a flow chart of a switch cabinet discharge fault detection method provided in an embodiment of the present application is shown;

[0015] Figure 2 A schematic diagram showing the change of the resistance of an ozone sensor provided by an embodiment of the present application as the ozone concentration increases;

[0016] Figure 3 A circuit diagram corresponding to a Holston bridge provided in an embodiment of the present application is shown;

[0017] Figure 4 A schematic flow chart of another switch cabinet discharge fault detection method provided in an embodiment of the present application is shown;

[0018] Figure 5 A schematic diagram showing a reading error of an analog-to-digital conversion module provided in an embodiment of the present application is shown;

[0019] Figure 6 A schematic diagram of the structure of a resistor box provided in an embodiment of the present application is shown;

[0020] Figure 7 A schematic diagram of one side of a target switch cabinet of a switch cabinet discharge fault detection system provided by an embodiment of the present application is shown;

[0021] Figure 8 A schematic diagram of a display screen provided in an embodiment of the present application is shown;

[0022] Fig. 9 A schematic structural diagram of a switch cabinet discharge fault detection system provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0024] In this embodiment, a switch cabinet discharge fault detection method is provided. Figure 1 As shown, the method includes:

[0025] Step 101, the analog-to-digital conversion module of the single-chip microcomputer reads the voltage of the Hothstone bridge and sends the read first voltage value to the cloud server, wherein the Hothstone bridge is arranged in the target switch cabinet, and the Hothstone bridge includes an ozone sensor. When a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes.

[0026] A switch cabinet discharge fault detection method provided in an embodiment of the present application can detect the discharge condition of a high-voltage switch cabinet in real time. Specifically, the present application is implemented by a single-chip microcomputer, a Hoston bridge, and a cloud server. Among them, the single-chip microcomputer is a microcontroller for receiving and processing the voltage output from the Hoston bridge; the Hoston bridge is an electronic measuring device installed in the target switch cabinet to monitor the change of ozone concentration inside the target switch cabinet. Specifically, the Hoston bridge includes an ozone sensor. When the target switch cabinet is discharged, ozone gas is usually generated. Therefore, by using the Hoston bridge including the ozone sensor, it is possible to indirectly determine whether a discharge fault has occurred in the target switch cabinet.

[0027] When a discharge fault occurs in the target switch cabinet, the ozone concentration in the target switch cabinet will increase due to the chemical reaction caused by the discharge, and the ozone sensor will react chemically with the ozone, causing the resistance of the ozone sensor to change. This resistance change can be reflected in the voltage output of the Holston bridge. Figure 2 As shown, the change of the resistance value of an ozone sensor with the increase of ozone concentration is shown. After that, the single-chip microcomputer reads the voltage of the Holston bridge through its built-in analog-to-digital converter (ADC) to obtain a first voltage value. The single-chip microcomputer then sends the first voltage value to the cloud server through a network (such as Wi-Fi, Ethernet, etc.) for further processing.

[0028] like Figure 3 As shown, a circuit corresponding to a Holston bridge is shown. Among them, resistors R1, R2, R0 and R S The Holston bridge is composed of a power supply V DD Power supply, resistor R0 is an adjustable resistor, resistor R S It is an ozone sensor.

[0029] Step 102: After receiving the first voltage value, the cloud server inputs the first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet based on the current ozone concentration value.

[0030] In this embodiment, after the cloud server receives the first voltage value sent by the single-chip microcomputer, it can input it into a preset ozone concentration recognition model, which is a trained algorithm or program that can calculate the current ozone concentration value in the target switch cabinet based on the input voltage value. After that, the cloud server can determine whether a discharge fault has occurred in the target switch cabinet based on the calculated current ozone concentration value. For example, the cloud server compares the calculated current ozone concentration value with the first preset concentration threshold. If the current ozone concentration value exceeds the first preset concentration threshold, the cloud server can determine that a discharge fault has occurred in the target switch cabinet and can trigger a series of response measures, such as sending an alarm to relevant personnel, recording fault information, and remotely cutting off the power supply; if the current ozone concentration value does not exceed the first preset concentration threshold, it is considered that the target switch cabinet is operating normally and no further measures are required.

[0031] By applying the technical solution of this embodiment, the single-chip microcomputer reads the voltage of the Hosstone bridge through its built-in analog-to-digital conversion module to obtain a first voltage value. The single-chip microcomputer then sends the first voltage value to the cloud server through the network for further processing. After receiving the first voltage value sent by the single-chip microcomputer, the cloud server can input it into a preset ozone concentration recognition model to calculate the current ozone concentration value in the target switch cabinet. Afterwards, the cloud server can determine whether a discharge fault has occurred in the target switch cabinet based on the calculated current ozone concentration value. The embodiment of the present application realizes real-time intelligent detection of discharge faults in the target switch cabinet by real-time monitoring of the changes in ozone concentration in the target switch cabinet and utilizing the collaborative work of the Hosstone bridge, the single-chip microcomputer, the cloud server and the ozone concentration recognition model; through low-cost devices such as the single-chip microcomputer and the Hosstone bridge with an ozone sensor, and using the cloud server as a computing terminal, the hardware cost is saved, the space occupied in the switch cabinet is reduced, and the complexity and cost of the monitoring system can be greatly reduced.

[0032] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another switch cabinet discharge fault detection method is provided, such as Figure 4 As shown, the method includes:

[0033] In step 201, the analog-to-digital conversion module of the single-chip microcomputer reads the analog voltage input by the high-precision digital power supply, and uses the read voltage value as the voltage prediction value, determines the actual voltage value corresponding to the analog voltage based on the high-precision digital power supply, constructs a first training sample according to the voltage prediction value and the actual voltage value, and sends the first training sample to the cloud server.

[0034] In this embodiment, the analog-to-digital conversion module of the single-chip microcomputer has a certain error initially, so it can be pre-calibrated before use. Figure 5 As shown in FIG. 1 , a schematic diagram of the difference between the controller reading voltage obtained by the analog-to-digital conversion module of the single-chip microcomputer (i.e., the controller in the figure) reading the input voltage of the high-precision digital power supply and the actual input voltage of the high-precision digital power supply is shown. Since the resistance of each resistor in the Hosston bridge is large, the output voltage of the Hosston bridge is small, so the small reading error of the analog-to-digital conversion module has a large impact on the final discharge fault judgment result. In order to improve the accuracy of discharge fault detection in the target switch cabinet, the reading of the analog-to-digital conversion module can be corrected. First, the analog-to-digital conversion module in the single-chip microcomputer reads the analog voltage input by the high-precision digital power supply, and uses the read voltage value as the voltage prediction value. The "prediction" here means that this is the voltage estimation value read by the analog-to-digital conversion module of the single-chip microcomputer, which can be affected by various factors such as the accuracy, nonlinearity, and noise of the analog-to-digital conversion module, resulting in differences from the true value of the voltage. The high-precision digital power supply can accurately know the true value of the analog voltage it outputs, that is, the above-mentioned true value of the voltage. Therefore, this true value of the voltage can be used as a reference and subsequently compared with the voltage prediction value read by the analog-to-digital conversion module of the single-chip microcomputer. Next, the MCU combines the predicted voltage value corresponding to the analog voltage with the true voltage value obtained through the high-precision digital power supply to form first training samples, and sends these first training samples to the cloud server. The powerful computing power and data processing capabilities of cloud computing are used to process these first training samples, avoiding resource limitations of complex calculations on the MCU.

[0035] For example, a high-precision digital power supply is used as the input voltage of the analog-to-digital conversion module. Assuming that the analog-to-digital conversion module of the single-chip microcomputer can read a voltage range of 0-3.3V, the voltage input to the analog-to-digital conversion module by the high-precision digital power supply is adjusted at intervals of 0.1V from 0V, and the analog-to-digital conversion module is set to read once per second, and 1 reading is read each time, that is, 1 voltage prediction value is read per second. After reading 100 voltage prediction values, the voltage input to the analog-to-digital conversion module by the high-precision digital power supply is adjusted at intervals of 0.1V, and read again until the maximum voltage of 3.3V is reached, and 3400 voltage prediction values ​​can be obtained. The single-chip microcomputer constructs 3400 first training samples based on the 3400 voltage prediction values ​​read by the analog-to-digital conversion module, and each first training sample includes a voltage prediction value and a voltage true value corresponding to the voltage prediction value. These 3400 first training samples are then sent to the cloud server.

[0036] In step 202, the cloud server trains the initial correction model using the first training sample until the model loss value of the initial correction model is less than a first preset loss threshold, thereby obtaining a voltage correction model corresponding to the analog-to-digital conversion module.

[0037] In this embodiment, after receiving the first training samples, the cloud server uses these first training samples to train the initial correction model, and finally obtains the voltage correction model. Specifically, the voltage prediction value in the first training sample can be used as the model input, and the model loss value is calculated according to the difference between the voltage real value in the first training sample and the model output. The goal of the voltage correction model is to learn how to adjust or correct the voltage prediction value read by the analog-to-digital conversion module to make it closer to the real value of the voltage. The training process is an iterative process, and the cloud server can continuously update the parameters of the initial correction model until the model loss value of the initial correction model is less than the first preset loss threshold. At this time, it indicates that the model has learned enough information and can effectively correct the error of the voltage value read by the analog-to-digital conversion module. When the training is completed, an optimized voltage correction model is obtained on the cloud server, which can give an output closer to the real value of the voltage according to the voltage prediction value read by the analog-to-digital conversion module. It should be noted that after obtaining the voltage correction model, the voltage correction model can be directly used to correct the voltage value read by the analog-to-digital conversion module. The embodiment of the present application achieves the effect of correcting the voltage value read by the analog-to-digital conversion module of the single-chip microcomputer by combining a high-precision digital power supply, a single-chip microcomputer and a cloud server, so that the corrected voltage value can be closer to the actual value of the voltage.

[0038] In step 203, the analog-to-digital conversion module of the single-chip microcomputer reads the voltage of the Hothstone bridge and sends the read first voltage value to the cloud server, wherein the Hothstone bridge is arranged in the target switch cabinet, and the Hothstone bridge includes an ozone sensor. When a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes.

[0039] Step 204: After receiving the first voltage value, the cloud server inputs the first voltage value into the voltage correction model to obtain a corrected first voltage value; inputs the corrected first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet based on the current ozone concentration value.

[0040] In this embodiment, when monitoring the discharge condition of the target switch cabinet, the cloud server can input the received first voltage value into the voltage correction model obtained after the above training is completed to obtain a corrected first voltage value, and then calculate the current ozone concentration value in the target switch cabinet according to the corrected first voltage value, so that the calculated current ozone concentration value can be closer to the actual value, thereby improving the accuracy of discharge detection in the target switch cabinet.

[0041] by Figure 3Taking the circuit corresponding to the Hoston bridge as an example, the analog-to-digital conversion module of the single-chip microcomputer can read a voltage range of 0-3.3V. The voltage output by the Hoston bridge is generally between 1-2V after being amplified 50 times by the operational amplifier, which is an ideal voltage reading range for the analog-to-digital conversion module. After the analog-to-digital conversion module built into the single-chip microcomputer reads the voltage, the single-chip microcomputer can send the read voltage to the WiFi module through the serial port in the form of a floating point number. The WiFi module works in the receive / transmit mode and sends the read voltage value to the cloud server. The cloud server brings the read voltage value into the voltage correction model and outputs the corrected voltage value.

[0042] In an embodiment of the present application, optionally, the Hoston bridge includes an adjustable resistor; before step 203, the method further includes: powering the Hoston bridge and the single-chip microcomputer by a preset battery, and initializing a first resistance value of the adjustable resistor in the Hoston bridge; the analog-to-digital conversion module of the single-chip microcomputer reads the Hoston bridge voltage, and sends the read second voltage value to the cloud server; after the cloud server receives the second voltage value, the second voltage value is input into the voltage correction model to obtain a corrected second voltage value, and when the corrected second voltage value is greater than a preset voltage threshold, the second resistance value of the adjustable resistor is calculated by a genetic algorithm, and the second resistance value is sent to the single-chip microcomputer; the single-chip microcomputer adjusts the adjustable resistor according to the received second resistance value so that the voltage value corresponding to the Hoston bridge voltage is less than or equal to the preset voltage threshold.

[0043] In this embodiment, before using the Hoston bridge to detect the discharge condition in the target switch cabinet, the Hoston bridge and the like may be initialized. First, a preset battery is used to power the Hoston bridge and the single-chip microcomputer, and at the same time, the adjustable resistor in the Hoston bridge is initialized to a preset first resistance value. Figure 3 The circuit of the Hoston bridge shown can be initialized by using a preset battery power supply to power the Hoston bridge, the amplifier and the single-chip microcomputer. After the single-chip microcomputer is started, it can perform a self-test to confirm whether it can communicate normally with the cloud server, and at the same time confirm whether the analog-to-digital conversion module can read the voltage value from the Hoston bridge. Specifically, the single-chip microcomputer can confirm whether it can communicate normally with the WiFi module through the serial port. If it can, it is considered that it can communicate normally with the cloud server. The first resistance value of the adjustable resistor can be set to 100 kilo-ohms by default, the initial magnification of the operational amplifier can be 50 times, and the adjustable resistor can adjust the resistance value in the range of 0-999999 ohms. The single-chip microcomputer of the embodiment of the present application communicates with the cloud server through the WiFi module, and there is no need to arrange communication lines and power lines on site, which is simple and convenient, and the system can be remotely maintained and updated; using batteries as a power supply method, there is no need to require additional power on site, which can reduce dependence on the on-site environment.

[0044] After the initialization is completed, the single-chip computer monitors the voltage on the Hoston bridge, and reads the Hoston bridge voltage through its built-in analog-to-digital conversion module to obtain a second voltage value, and sends the obtained second voltage value to the cloud server. After receiving the second voltage value, the cloud server inputs it into the previously trained voltage correction model. The voltage correction model outputs a corrected second voltage value based on the input second voltage value. If the corrected second voltage value is greater than the preset voltage threshold, it means that the current output voltage of the Hoston bridge does not meet the requirements and needs to be adjusted. For example, the preset voltage threshold is 0.001V. If the corrected second voltage value is greater than 0.001V, it means that the current output voltage of the Hoston bridge does not meet the requirement of being close to zero. Subsequently, the cloud server can use a genetic algorithm to calculate the second resistance value of the adjustable resistor, and send the calculated second resistance value back to the single-chip computer. Here, the second resistance value is the resistance value of the adjustable resistor that makes the Hoston bridge voltage less than the preset voltage threshold.

[0045] After the single-chip microcomputer receives the second resistance value sent by the cloud server, the single-chip microcomputer adjusts the adjustable resistor in the Hoston bridge to the second resistance value through the corresponding control interface. The resistance value of the adjusted adjustable resistor will change the output voltage of the Hoston bridge so that its corresponding voltage value is less than or equal to the preset voltage threshold. After initializing the single-chip microcomputer, etc., the embodiment of the present application adjusts the resistance value of the adjustable resistor in the Hoston bridge so that the Hoston bridge voltage is close to zero. When the resistance value of the ozone sensor changes later, the Hoston bridge voltage changes immediately and is no longer a value close to zero, thereby facilitating the subsequent calculation of the ozone concentration value.

[0046] In an embodiment of the present application, optionally, the adjustable resistor is a resistor box, which includes a plurality of multiplexers, and the output resistance value of each of the multiplexers corresponds to the resistance value of a digit of the resistor box; the "calculating the second resistance value of the adjustable resistor by a genetic algorithm" includes: the cloud server randomly generates a preset number of code sets by a genetic algorithm, and sends the preset number of code sets to the single-chip microcomputer, wherein the number of codes included in each of the code sets is the same as the total number of control terminals of each of the multiplexers, each code is used to indicate the connection state of the corresponding control terminal, and the control terminal of each multiplexer is used to control the output resistance value of the multiplexer; the single-chip microcomputer adjusts the connection state of each control terminal in the resistor box according to the code included in each code set, and passes the code set to the single-chip microcomputer after the adjustment. The analog-to-digital conversion module reads the Hosstone bridge voltage corresponding to each coding set, and sends the read third voltage value to the cloud server; after receiving a preset number of third voltage values, the cloud server inputs each of the third voltage values ​​into the voltage correction model to obtain a corrected third voltage value, determines a target voltage value from a preset number of corrected third voltage values ​​according to a preset membership function, adjusts the coding values ​​in the preset number of coding sets through a genetic algorithm to obtain an updated coding set, and returns to the step of sending the preset number of coding sets to the single-chip microcomputer until the preset conditions are met, determines a target voltage value with the smallest absolute value from multiple target voltage values, and uses the resistance value indicated by the coding set corresponding to the target voltage value with the smallest absolute value as the second resistance value.

[0047] In this embodiment, the adjustable resistor in the Hosstone bridge can be specifically a resistor box, which is composed of a plurality of multiplexers, each of which controls the resistance value of a digit in the resistor box. By changing the connection state of the control end of the multiplexer, different resistors can be selected to access the circuit, thereby changing the resistance value on the digit. For example, the resistor box includes 6 multiplexers, namely multiplexers 1 to multiplexers 6, then the resistance value of multiplexer 1 can be the resistance value on the unit digit of the resistor box, the resistance value of multiplexer 2 can be the resistance value on the tens digit of the resistor box, the resistance value of multiplexer 3 can be the resistance value on the hundreds digit of the resistor box, and so on, then the resistor boxes corresponding to the 6 multiplexers can adjust the resistance value within the range of 0-999999 ohms.

[0048] When calculating the second resistance value of the adjustable resistor through a genetic algorithm, first, the cloud server can randomly generate a preset number of code sets. Among them, the number of codes in each code set is the same as the total number of control terminals of the multiplexer, and each code represents the connection status of a control terminal (such as on / off). Assuming that each multiplexer corresponds to 4 control terminals, and the resistor box includes 6 multiplexers, then there are a total of 24 control terminals. At this time, each code set can be set to include 24 codes, and each code can have two values ​​0 or 1, where 0 can represent off and 1 can represent on. Then, through each code set, the connection status of each control terminal in different multiplexers in the resistor box can be known. Afterwards, these code sets can be sent to the microcontroller.

[0049] The single-chip microcomputer adjusts the control terminal connection state of each multiplexer in the resistor box according to the received code set. Specifically, the single-chip microcomputer can use each code set to adjust the resistor box once, that is, according to the received preset number of code sets, the resistor box can be adjusted a preset number of times. After each adjustment, the single-chip microcomputer reads the voltage value of the Holston bridge, that is, the third voltage value, through the analog-to-digital conversion module, and sends these third voltage values ​​back to the cloud server.

[0050] After receiving all the third voltage values, the cloud server uses the voltage correction model to correct these values ​​to improve the accuracy, and then a preset number of corrected third voltage values ​​can be obtained. Afterwards, according to the preset membership function, a target voltage value is determined from the corrected third voltage values, and the target voltage value is the voltage value closest to the ideal value. Here, the membership function is a function used to evaluate the degree of proximity between each third voltage value and the final desired voltage value. Then, the cloud server can use a genetic algorithm to iteratively optimize these coding sets. In each iteration, the coding values ​​in the coding set are adjusted according to the quality of the voltage value (that is, the control end connection state of the multiplexer is changed) to generate a new coding set. And the new coding set is sent to the single-chip microcomputer again, and the single-chip microcomputer adjusts the adjustable resistance according to the new coding set... This process is repeated until the preset conditions are met. Here, the preset condition can be reaching the maximum number of iterations, etc. From the target voltage values ​​obtained after multiple iterations, the coding set corresponding to the target voltage value with the smallest absolute value is selected, and the resistance value indicated by this coding set is the second resistance value required, that is, the resistance box setting that can minimize the Holston bridge voltage.

[0051] For example, Figure 6As shown, the resistor box includes 6 multiplexers, each of which has 4 control terminals (A, B, C, D, etc. in the figure). The cloud server uses a genetic algorithm to randomly generate 50 coding sets, each of which includes 24 binary codes, and each binary code corresponds to the connection status of a control terminal. Then each coding set containing 24 binary codes is sent to the single-chip microcomputer at one time through the wireless module. The single-chip microcomputer controls the resistor box according to the coding set, reads the voltage value corresponding to each coding set, and feeds it back to the cloud server through the communication module. The cloud server brings the voltage value into the voltage correction model, and then substitutes the corrected voltage value into the fitness function. Finally, a target voltage value can be determined from the 50 corrected voltage values. Continue to use the genetic algorithm to iterate 500 times to get 500 target voltage values, and then determine the second resistance value based on these 500 target voltage values.

[0052] It should be noted that the signal at the control end can be represented by 0 and 1, 0 represents a low level, and 1 represents a high level. Each multiplexer is determined by 4 control terminals to determine the conduction route of the multiplexer, and 4 binary can control up to 16 signal paths. In this application, each multiplexer has only 10 resistors, so the binary encoding method of the genetic algorithm is used to encode the 4 control terminals of the 6 multiplexers. Since the sensor resistance value is very large, decimal resistors are not included. The upper limit of the output resistance of each multiplexer is 9. The 4 control terminals of each multiplexer are converted into decimal resistance values, representing units, tens, hundreds, thousands, ten thousand, and one hundred thousand respectively. The absolute value of the Hoston bridge voltage corresponding to each coding set is used as the fitness function, and the minimum value of the function is the target value required. Since the final desired voltage value in this application is zero, the membership function of this application can be specifically recorded as:

[0053] ;

[0054] in, V O is the Holston bridge voltage.

[0055] The embodiment of the present application uses a genetic algorithm to form a closed-loop control system for the entire process, in which the single-chip microcomputer is responsible for data acquisition and hardware adjustment, and the cloud server is responsible for data processing and feedback calculation. By continuously iterating this process, the output voltage of the Hosston bridge can be automatically and accurately controlled to meet the demand for the output voltage to be close to zero. In traditional zero-position correction, an open-loop system calculated by a theoretical formula is used for correction. The value calculated by the formula does not consider factors such as line loss, component errors and interference, so it often brings large errors. Compared with traditional PID (Proportional-Integral-Derivative) control, the zero-position correction system of the embodiment of the present application does not need to model the entire system, nor does it need to solve PID parameters, which reduces the debugging time before the deployment of the detection system. In addition, the embodiment of the present application combines the powerful capabilities of cloud computing and the real-time control capabilities of the single-chip microcomputer to achieve remote monitoring and precise control of complex systems.

[0056] In an embodiment of the present application, optionally, after the "single-chip microcomputer adjusts the adjustable resistor according to the received second resistance value", the method also includes: the analog-to-digital conversion module of the single-chip microcomputer reads the Hosstone bridge voltage under different preset ozone concentration values, and sends the read fourth voltage value to the cloud server; after the cloud server receives the fourth voltage value, the fourth voltage value is input into the voltage correction model to obtain a corrected fourth voltage value, and a second training sample is constructed based on the corrected fourth voltage value and the corresponding preset ozone concentration value, and the initial ozone concentration recognition model is trained by the second training sample until the model loss value of the initial ozone concentration recognition model is less than the second preset loss threshold, so as to obtain the preset ozone concentration recognition model.

[0057] In this embodiment, after the single-chip microcomputer adjusts the adjustable resistor according to the second resistance value, the single-chip microcomputer reads the voltage output of the Hosstone bridge at different preset ozone concentration values ​​through its built-in analog-to-digital conversion module, and sends the read fourth voltage value to the cloud server. Among them, the fourth voltage value reflects the response of the Hosstone bridge to different ozone concentrations. After receiving the fourth voltage value sent by the single-chip microcomputer, the cloud server inputs the fourth voltage value into the voltage correction model to obtain the corrected fourth voltage value. After that, each corrected fourth voltage value can be combined with the corresponding preset ozone concentration value to construct a second training sample, which contains all the information required for the initial ozone concentration recognition model training, that is, the input (corrected fourth voltage value) and the output (preset ozone concentration value). The initial ozone concentration recognition model set in the cloud server can be constructed based on a machine learning algorithm. The initial ozone concentration recognition model does not have a good recognition ability, and its parameters need to be optimized through training. Therefore, the cloud server uses the constructed second training sample to train the initial ozone concentration recognition model. During the training process, the model continuously adjusts its internal parameters to minimize the difference between the model output and the actual ozone concentration value (i.e., the model loss value). When the model loss value is less than the second preset loss threshold, the training process ends and a trained preset ozone concentration recognition model is obtained. The trained preset ozone concentration recognition model can be used to monitor and identify the ozone concentration in the switch cabinet environment in real time. When the single-chip microcomputer reads the voltage value of the Hoston bridge again, it can be input into the model after being corrected by the voltage correction model to obtain the estimated value of the ozone concentration in the current switch cabinet environment. The embodiment of the present application realizes the training of the initial ozone concentration recognition model by combining a single-chip microcomputer, a cloud server, an analog-to-digital conversion module, a voltage correction model, and a machine learning algorithm, and finally obtains a preset ozone concentration recognition model that can accurately identify the ozone concentration. The relationship between the Hoston bridge voltage and the ozone concentration in the switch cabinet can be established, thereby improving the recognition accuracy and efficiency of the ozone concentration.

[0058] In an embodiment of the present application, optionally, after step 204, the method further includes: when the current ozone concentration value is greater than a first preset concentration threshold, determining that a discharge fault occurs in the target switch cabinet; accordingly, after "determining that a discharge fault occurs in the target switch cabinet", the method further includes: when the current ozone concentration value is greater than a second preset concentration threshold, generating an end test instruction, and sending the end test instruction to the single-chip microcomputer, wherein the second preset concentration threshold is greater than the first preset concentration threshold; when the current ozone concentration value is less than or equal to the second preset concentration threshold, generating a continue test instruction, and starting timing, continuing to calculate the ozone concentration value until a preset time is reached, and determining whether the ozone concentration value within the preset time increases; when it increases, determining that a continuous discharge fault occurs in the target switch cabinet, and generating an end test instruction, otherwise, determining that a short-time discharge fault occurs in the target switch cabinet, and generating a continue test instruction.

[0059] In this embodiment, first, the calculated current ozone concentration value can be compared with the first preset concentration threshold. When the current ozone concentration value is greater than the first preset concentration threshold, it is determined that a discharge fault has occurred in the target switch cabinet. Here, the first preset concentration threshold can be a lower concentration value set based on experience, which is used to preliminarily determine the existence of the discharge phenomenon.

[0060] If it is determined that a discharge fault has occurred in the target switch cabinet, the magnitude relationship between the current ozone concentration value and the second preset concentration threshold can be further determined. If the current ozone concentration value is further greater than the second preset concentration threshold, it means that the discharge phenomenon in the target switch cabinet is relatively serious, and it is not meaningful to continue the detection. Therefore, an end test instruction can be generated and sent to the single-chip microcomputer. After receiving the end test instruction, the single-chip microcomputer can stop reading the Holston bridge voltage. Here, the second preset concentration threshold is greater than the first preset concentration threshold. In addition, the test here refers to a test for detecting the discharge situation in the target switch cabinet. If the current ozone concentration value is less than or equal to the second preset concentration threshold, it means that although it is confirmed that the target switch cabinet has a discharge fault, the discharge is not serious. At this time, a continue test instruction can be generated, and the timing can be started to further observe the change of ozone concentration in order to distinguish the type of discharge fault. Within the preset time, the ozone concentration value can continue to be calculated. If the ozone concentration value shows an increasing trend during this period of time, it means that the discharge phenomenon is continuing, so it is determined that a continuous discharge fault has occurred in the target switch cabinet. At this time, the end test instruction is generated, and an alarm or other emergency measures can also be triggered. On the contrary, if the ozone concentration value does not increase or increases insignificantly within the preset time, it is judged as a short-term discharge fault. This fault may be instantaneous or sporadic, and has little impact on the long-term operation of the equipment. At this time, a continue test instruction can be generated, indicating that the test can continue. The embodiment of the present application not only realizes the timely detection of the switch cabinet discharge fault by monitoring the change of ozone concentration, but also further judges the type of discharge fault by the change trend of ozone concentration, thereby providing more detailed information for subsequent maintenance and processing. In addition, by sending test instructions to the single-chip microcomputer, the test process can be flexibly controlled.

[0061] After determining that there is a discharge fault in the switch cabinet, the cloud server can also send fault information to the preset terminal. The fault information may include the switch cabinet identification of the switch cabinet where the discharge fault occurs, the time of discharge occurrence, etc., to remind the maintenance personnel corresponding to the preset terminal to handle the discharge fault in time. It is also possible to generate fault information and send it to the preset terminal after determining the type of discharge fault, so as to help the maintenance personnel understand the type of discharge fault.

[0062] In an embodiment of the present application, optionally, after the "generating an end test instruction", the method further includes: generating a power-off instruction for the target switch cabinet, and sending the power-off instruction to a power control device of the target switch cabinet.

[0063] In this embodiment, after generating the end test instruction, in order to prevent the discharge fault in the switch cabinet from further expanding or causing more serious consequences, the cloud server can generate a power-off instruction. This power-off instruction can be sent to the power control device of the target switch cabinet. The power control device is a device responsible for controlling the power on and off of the switch cabinet. When the power control device receives the power-off instruction, it can perform corresponding operations to cut off the power supply of the target switch cabinet. In this way, all electrical equipment in the target switch cabinet can stop working, thereby avoiding further discharge or other potential hazards. The embodiment of the present application reduces the risks of fire, explosion, etc. caused by switch cabinet failure by controlling the target switch cabinet to automatically cut off the power supply when a serious discharge fault is detected, significantly improves the safety and reliability of electrical equipment such as switch cabinets, and protects personnel safety. In addition, through remote monitoring and instruction issuance by the cloud server, simultaneous management and control of multiple switch cabinets can be achieved, improving operation and maintenance efficiency and management level.

[0064] In the embodiment of the present application, optionally, a display screen and a memory may be provided on one side of the target switch cabinet. Figure 7 As shown, the central controller can be a single-chip microcomputer, the communication module can be the above-mentioned WiFi module, and the memory can be used to store data for the single-chip microcomputer, such as storing the read voltage value, etc. A display screen is set on one side of the target switch cabinet, so that the user can better understand the operating status of the switch cabinet discharge fault detection system. Figure 8 As shown, a display interface of a display screen is shown. The display interface includes a power indicator light, a network status light, a system status light, etc. The power indicator light can be used to show whether the power connection is abnormal, the network status light can be used to show whether the network status is abnormal, and the system status light can be used to show whether the system status of the switch cabinet discharge fault detection system is abnormal.

[0065] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides a switch cabinet discharge fault detection system, such as Fig. 9 As shown, the system includes:

[0066] A single-chip computer, used for reading a voltage of a Hothstone bridge through an analog-to-digital conversion module, and sending the read first voltage value to a cloud server, wherein the Hothstone bridge is arranged in a target switch cabinet, and the Hothstone bridge includes an ozone sensor, and when a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes;

[0067] After receiving the first voltage value, the cloud server inputs the first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet based on the current ozone concentration value.

[0068] Optionally, the single-chip computer is further used to read the analog voltage of the high-precision digital power supply input through the analog-to-digital conversion module before reading the Holston bridge voltage through the analog-to-digital conversion module, and use the read voltage value as the voltage prediction value, determine the voltage true value corresponding to the analog voltage based on the high-precision digital power supply, construct a first training sample according to the voltage prediction value and the voltage true value, and send the first training sample to the cloud server;

[0069] The cloud server is further used to train the initial correction model through the first training sample until the model loss value of the initial correction model is less than a first preset loss threshold, thereby obtaining a voltage correction model corresponding to the analog-to-digital conversion module;

[0070] Correspondingly, the cloud server is further used to input the first voltage value into the voltage correction model after receiving the first voltage value, so as to obtain a corrected first voltage value.

[0071] Optionally, the Holston bridge includes an adjustable resistor; and the system further includes:

[0072] An initialization module, used for supplying power to the Holston bridge and the single-chip microcomputer through a preset battery and initializing a first resistance value of an adjustable resistor in the Holston bridge before the single-chip microcomputer reads the Holston bridge voltage through the analog-to-digital conversion module;

[0073] The single chip microcomputer is further used to read the Holston bridge voltage through the analog-to-digital conversion module, and send the read second voltage value to the cloud server;

[0074] The cloud server is further configured to, after receiving the second voltage value, input the second voltage value into the voltage correction model to obtain a corrected second voltage value, and when the corrected second voltage value is greater than a preset voltage threshold, calculate the second resistance value of the adjustable resistor by a genetic algorithm, and send the second resistance value to the single chip computer;

[0075] The single chip microcomputer is further used to adjust the adjustable resistor according to the received second resistance value so that the voltage value corresponding to the Holston bridge voltage is less than or equal to the preset voltage threshold.

[0076] Optionally, the adjustable resistor is a resistor box, the resistor box includes a plurality of multiplexers, and the output resistance value of each of the multiplexers corresponds to a resistance value on a digit of the resistor box;

[0077] The cloud server is further used to randomly generate a preset number of code sets through a genetic algorithm, and send the preset number of code sets to the single-chip microcomputer, wherein the number of codes included in each of the code sets is the same as the total number of control terminals of each of the multiplexers, each code is used to indicate the connection state of the corresponding control terminal, and the control terminal of each multiplexer is used to control the output resistance of the multiplexer;

[0078] The single chip microcomputer is further used to adjust the connection state of each control terminal in the resistance box according to the code included in each code set, and read the Hosstone bridge voltage corresponding to each code set through the analog-to-digital conversion module after the adjustment, and send the read third voltage value to the cloud server;

[0079] The cloud server is also used to, after receiving a preset number of third voltage values, input each of the third voltage values ​​into the voltage correction model to obtain a corrected third voltage value, determine a target voltage value from the preset number of corrected third voltage values ​​according to a preset membership function, adjust the code values ​​in the preset number of code sets through a genetic algorithm to obtain an updated code set, and return to the step of sending the preset number of code sets to the single-chip computer until a preset condition is met, determine the target voltage value with the smallest absolute value from multiple target voltage values, and use the resistance value indicated by the code set corresponding to the target voltage value with the smallest absolute value as the second resistance value.

[0080] Optionally, the single chip microcomputer is further used to adjust the adjustable resistor according to the received second resistance value, read the Holston bridge voltage under different preset ozone concentration values ​​through the analog-to-digital conversion module, and send the read fourth voltage value to the cloud server;

[0081] The cloud server is further used to, after receiving the fourth voltage value, input the fourth voltage value into the voltage correction model to obtain a corrected fourth voltage value, and construct a second training sample based on the corrected fourth voltage value and the corresponding preset ozone concentration value, and train the initial ozone concentration recognition model through the second training sample until the model loss value of the initial ozone concentration recognition model is less than a second preset loss threshold, thereby obtaining the preset ozone concentration recognition model.

[0082] Optionally, the cloud server is further used for:

[0083] When the current ozone concentration value is greater than a first preset concentration threshold, determining that a discharge fault occurs in the target switch cabinet;

[0084] Accordingly, the cloud server is also used for:

[0085] After determining that the target switch cabinet has a discharge fault, when the current ozone concentration value is greater than a second preset concentration threshold, generating an end test instruction, and sending the end test instruction to the single chip microcomputer, wherein the second preset concentration threshold is greater than the first preset concentration threshold;

[0086] When the current ozone concentration value is less than or equal to the second preset concentration threshold, a continue test instruction is generated, and the timing is started. The ozone concentration value is continuously calculated until the preset time is reached, and it is determined whether the ozone concentration value within the preset time increases; when it increases, it is determined that a continuous discharge fault occurs in the target switch cabinet, and an end test instruction is generated; otherwise, it is determined that a short-time discharge fault occurs in the target switch cabinet, and a continue test instruction is generated.

[0087] Optionally, the cloud server is further used for:

[0088] After the test end instruction is generated, a power off instruction of the target switch cabinet is generated, and the power off instruction is sent to the power control device of the target switch cabinet.

[0089] It should be noted that for other corresponding descriptions of the functional units involved in the switch cabinet discharge fault detection system provided in the embodiment of the present application, reference can be made to Figures 1 to 8 The corresponding description in the method will not be repeated here.

[0090] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0091] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A switch cabinet discharge fault detection method, characterized in that: include: The analog-to-digital conversion module of the single-chip microcomputer reads the voltage of the Hothstone bridge and sends the read first voltage value to the cloud server, wherein the Hothstone bridge is arranged in the target switch cabinet, and the Hothstone bridge includes an ozone sensor. When a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes; After receiving the first voltage value, the cloud server inputs the first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determines whether a discharge fault occurs in the target switch cabinet according to the current ozone concentration value; Before the analog-to-digital conversion module of the single-chip microcomputer reads the Holston bridge voltage, the method further includes: The analog-to-digital conversion module of the single-chip microcomputer reads the analog voltage input by the high-precision digital power supply, and uses the read voltage value as the voltage prediction value, determines the voltage real value corresponding to the analog voltage based on the high-precision digital power supply, constructs a first training sample according to the voltage prediction value and the voltage real value, and sends the first training sample to the cloud server; The cloud server trains the initial correction model through the first training sample until the model loss value of the initial correction model is less than the first preset loss threshold, thereby obtaining a voltage correction model corresponding to the analog-to-digital conversion module; Correspondingly, after the cloud server receives the first voltage value, the method further includes: The cloud server inputs the first voltage value into the voltage correction model to obtain a corrected first voltage value; The Hothstone bridge includes an adjustable resistor; before the analog-to-digital conversion module of the single-chip computer reads the Hothstone bridge voltage, the method further includes: Powering the Holston bridge and the single-chip microcomputer via a preset battery, and initializing a first resistance value of an adjustable resistor in the Holston bridge; The analog-to-digital conversion module of the single-chip microcomputer reads the Holston bridge voltage and sends the read second voltage value to the cloud server; After receiving the second voltage value, the cloud server inputs the second voltage value into the voltage correction model to obtain a corrected second voltage value. When the corrected second voltage value is greater than a preset voltage threshold, the second resistance value of the adjustable resistor is calculated by a genetic algorithm, and the second resistance value is sent to the single chip computer. The single chip microcomputer adjusts the adjustable resistor according to the received second resistance value so that the voltage value corresponding to the Holston bridge voltage is less than or equal to the preset voltage threshold; The adjustable resistor is a resistor box, which includes a plurality of multiplexers, and the output resistance value of each multiplexer corresponds to a resistance value of a digit of the resistor box; the second resistance value of the adjustable resistor is calculated by a genetic algorithm, including: The cloud server randomly generates a preset number of code sets through a genetic algorithm, and sends the preset number of code sets to the single-chip microcomputer, wherein the number of codes included in each code set is the same as the total number of control terminals of each multiplexer, each code is used to indicate the connection state of the corresponding control terminal, and the control terminal of each multiplexer is used to control the output resistance value of the multiplexer; The single chip microcomputer adjusts the connection state of each control terminal in the resistance box according to the code included in each code set, and reads the Hosstone bridge voltage corresponding to each code set through the analog-to-digital conversion module after the adjustment, and sends the read third voltage value to the cloud server; After receiving a preset number of third voltage values, the cloud server inputs each of the third voltage values ​​into the voltage correction model to obtain a corrected third voltage value, determines a target voltage value from the preset number of corrected third voltage values ​​according to a preset membership function, adjusts the code values ​​in the preset number of code sets through a genetic algorithm to obtain an updated code set, and returns to the step of sending the preset number of code sets to the single-chip computer until a preset condition is met, determines a target voltage value with the smallest absolute value from multiple target voltage values, and uses the resistance value indicated by the code set corresponding to the target voltage value with the smallest absolute value as the second resistance value.

2. The method according to claim 1, characterized in that After the single chip microcomputer adjusts the adjustable resistor according to the received second resistance value, the method further includes: The analog-to-digital conversion module of the single-chip microcomputer reads the Holston bridge voltage under different preset ozone concentration values, and sends the read fourth voltage value to the cloud server; After receiving the fourth voltage value, the cloud server inputs the fourth voltage value into the voltage correction model to obtain a corrected fourth voltage value, and constructs a second training sample based on the corrected fourth voltage value and the corresponding preset ozone concentration value. The initial ozone concentration recognition model is trained using the second training sample until the model loss value of the initial ozone concentration recognition model is less than a second preset loss threshold, thereby obtaining the preset ozone concentration recognition model.

3. The method according to claim 1, characterized in that The determining, according to the current ozone concentration value, whether a discharge fault occurs in the target switch cabinet includes: When the current ozone concentration value is greater than a first preset concentration threshold, determining that a discharge fault occurs in the target switch cabinet; Correspondingly, after determining that a discharge fault occurs in the target switch cabinet, the method further includes: When the current ozone concentration value is greater than a second preset concentration threshold, generating an end test instruction and sending the end test instruction to the single chip microcomputer, wherein the second preset concentration threshold is greater than the first preset concentration threshold; When the current ozone concentration value is less than or equal to the second preset concentration threshold, a continue test instruction is generated, and the timing is started. The ozone concentration value is continuously calculated until the preset time is reached, and it is determined whether the ozone concentration value within the preset time increases; when it increases, it is determined that a continuous discharge fault occurs in the target switch cabinet, and an end test instruction is generated; otherwise, it is determined that a short-time discharge fault occurs in the target switch cabinet, and a continue test instruction is generated.

4. The method according to claim 3, characterized in that: After generating the end test instruction, the method further includes: A power-off instruction of the target switch cabinet is generated, and the power-off instruction is sent to a power control device of the target switch cabinet.

5. A switch cabinet discharge fault detection system, characterized in that: include: A single-chip computer, used for reading a voltage of a Hothstone bridge through an analog-to-digital conversion module, and sending the read first voltage value to a cloud server, wherein the Hothstone bridge is arranged in a target switch cabinet, and the Hothstone bridge includes an ozone sensor, and when a discharge fault occurs in the target switch cabinet, the resistance value of the ozone sensor changes; The cloud server is used for, after receiving the first voltage value, inputting the first voltage value into a preset ozone concentration identification model to obtain a current ozone concentration value in the target switch cabinet, and determining whether a discharge fault occurs in the target switch cabinet according to the current ozone concentration value; The single-chip computer is further used to read the analog voltage input by the high-precision digital power supply through the analog-to-digital conversion module before reading the Holston bridge voltage through the analog-to-digital conversion module, and use the read voltage value as the voltage prediction value, determine the voltage true value corresponding to the analog voltage based on the high-precision digital power supply, construct a first training sample according to the voltage prediction value and the voltage true value, and send the first training sample to the cloud server; The cloud server is further used to train the initial correction model through the first training sample until the model loss value of the initial correction model is less than a first preset loss threshold, thereby obtaining a voltage correction model corresponding to the analog-to-digital conversion module; Correspondingly, the cloud server is further configured to, after receiving the first voltage value, input the first voltage value into the voltage correction model to obtain a corrected first voltage value; The Holston bridge includes an adjustable resistor; the system also includes: An initialization module, used for supplying power to the Holston bridge and the single-chip microcomputer through a preset battery and initializing a first resistance value of an adjustable resistor in the Holston bridge before the single-chip microcomputer reads the Holston bridge voltage through the analog-to-digital conversion module; The single chip microcomputer is further used to read the Holston bridge voltage through the analog-to-digital conversion module, and send the read second voltage value to the cloud server; The cloud server is further configured to, after receiving the second voltage value, input the second voltage value into the voltage correction model to obtain a corrected second voltage value, and when the corrected second voltage value is greater than a preset voltage threshold, calculate the second resistance value of the adjustable resistor by a genetic algorithm, and send the second resistance value to the single chip computer; The single chip computer is further used to adjust the adjustable resistor according to the received second resistance value so that the voltage value corresponding to the Holston bridge voltage is less than or equal to the preset voltage threshold; The adjustable resistor is a resistor box, which includes a plurality of multiplexers, and the output resistance value of each multiplexer corresponds to the resistance value of a digit of the resistor box; The cloud server is further used to randomly generate a preset number of code sets through a genetic algorithm, and send the preset number of code sets to the single-chip microcomputer, wherein the number of codes included in each of the code sets is the same as the total number of control terminals of each of the multiplexers, each code is used to indicate the connection state of the corresponding control terminal, and the control terminal of each multiplexer is used to control the output resistance of the multiplexer; The single chip microcomputer is further used to adjust the connection state of each control terminal in the resistance box according to the code included in each code set, and read the Hosstone bridge voltage corresponding to each code set through the analog-to-digital conversion module after the adjustment, and send the read third voltage value to the cloud server; The cloud server is also used to, after receiving a preset number of third voltage values, input each of the third voltage values ​​into the voltage correction model to obtain a corrected third voltage value, determine a target voltage value from the preset number of corrected third voltage values ​​according to a preset membership function, adjust the code values ​​in the preset number of code sets through a genetic algorithm to obtain an updated code set, and return to the step of sending the preset number of code sets to the single-chip computer until a preset condition is met, determine the target voltage value with the smallest absolute value from multiple target voltage values, and use the resistance value indicated by the code set corresponding to the target voltage value with the smallest absolute value as the second resistance value.

Images