A high-voltage line detection system and method based on RISC-V

Abstract

The present invention relates to the field of detection technology, and in particular to a RISC-V-based high-voltage line detection system and method. The present invention automatically or manually controls the movement of a RISC-V-based terminal detection device on a high-voltage line to detect the real-time current at each point on the high-voltage line and collect a real-time image of the high-voltage line, thereby avoiding contact between maintenance personnel and the high-voltage line. The detected real-time current is sent to a cloud platform through a communication module. The cloud platform performs calculations based on the real-time current and issues an alarm based on the calculation results, so that maintenance personnel can handle the problem in time, thereby further improving work efficiency. When a terminal monitoring device displays a real-time image and identifies an obstacle based on the real-time image, an alarm is controlled to sound an alarm, and the RISC-V-based terminal detection device is automatically controlled to stop running, thereby reminding maintenance personnel to handle the obstacle as soon as possible, thereby helping to improve efficiency and avoiding the occurrence of dangerous situations.

Description

Technical Field

[0001] The present invention relates to the field of detection technology, and in particular to a high-voltage line detection system and method based on RISC-V. Background Art

[0002] High-voltage transmission lines are the lifeline of modern industrial society and national life. Whether they are normal or not is related to national livelihood and social stability. However, the long-distance transmission method and special environment will inevitably cause line failures, resulting in great loss of life and property. At the same time, in order to improve the economic benefits of the power grid, in recent years, power grid companies have invested a lot of energy in equipment, electricity technology, etc., aiming to improve the economy of distribution networks and reduce losses. However, the distribution network losses in some areas remain high, and even account for a very high proportion, causing serious losses to power grid companies. Analyzing the causes of this phenomenon, electricity theft has become the biggest obstacle to energy saving and consumption reduction.

[0003] In the prior art, manual inspection is mainly used to detect electricity theft through on-site contact detection. In addition, manual inspection is also mainly used to identify obstacles on power transmission lines. Telescopes are used to inspect whether there are foreign objects such as plastic bags and kites entangled in the lines, and paper materials are used to record them. This method that relies on manual inspection is not only inefficient but also has a certain risk of electric shock. Summary of the invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a high-voltage line detection system and method based on RISC-V, which helps to realize contactless detection and maintenance of high-voltage lines, thereby avoiding the risk of electric shock to a certain extent and helping to improve efficiency.

[0005] In order to implement the above technical solution, in a first aspect, the present invention provides a RISC-V-based high-voltage line detection system, including: a RISC-V-based terminal detection device, a communication module, a terminal monitoring device and a cloud platform;

[0006] RISC-V-based terminal detection equipment is set on the high-voltage line to detect the real-time current on the high-voltage line and collect real-time images on the high-voltage line;

[0007] The terminal monitoring device communicates with the RISC-V-based terminal detection device through a communication module, is used to obtain images collected by the RISC-V-based terminal detection device for real-time display, and recognize obstacles on the high-voltage line based on the acquired images;

[0008] The cloud platform is connected to the RISC-V-based terminal detection equipment and alarm through the communication module to obtain the real-time current on the high-voltage line, and calculate the line loss value of the high-voltage line based on the obtained real-time current;

[0009] The terminal monitoring device is also used to send a start signal and a stop signal to the RISC-V based terminal detection device.

[0010] Furthermore, the terminal monitoring device includes:

[0011] A display unit, used for displaying in real time the real-time image acquired by the RISC-V-based terminal detection device;

[0012] An image feature extraction unit, used to extract features from an image acquired by a RISC-V-based terminal detection device to generate a feature map;

[0013] The obstacle recognition unit uses the SPP module in the YOLOv4 detection algorithm framework to identify the feature map to identify obstacles. The SPP module includes three maximum pooling of different scales of 1*1, 5*5 and 13*13, and a 9*9 average pooling.

[0014] Further, the RISC-V-based terminal detection device includes: a real-time current detection unit;

[0015] The real-time current detection unit adopts a split structure including an upper device and a lower device;

[0016] The upper part of the equipment includes rotatable tires, motors and RISC-V-based microcontroller modules;

[0017] The rotatable tire is in contact with the high-voltage line; the RISC-V-based microcontroller module is connected to the terminal monitoring device through the communication module to receive the stop and start signals sent by the terminal monitoring device;

[0018] The RISC-V-based microcontroller module is connected to the tire through the motor to control the operation and stop of the motor based on the received stop and start signals, thereby driving the tire to move forward and stop on the high-voltage line;

[0019] The upper device and the lower device are connected by an automatic lock, and after the automatic lock is connected, an open-close mutual inductance coil is formed, so that the high-voltage line passes through the coil to detect the real-time current in the high-voltage line;

[0020] The RISC-V-based microcontroller module is connected to the open-and-close mutual inductance coil to obtain the detected real-time current and send it to the cloud platform through the communication module.

[0021] Furthermore, the communication module adopts a 4G wireless communication module.

[0022] On the other hand, the present invention provides a RISC-V-based high-voltage line detection method according to the RISC-V-based high-voltage line detection device described above, comprising the following steps:

[0023] Step 1: Detect the real-time current on the high-voltage line and collect the real-time image of the high-voltage line;

[0024] Step 2: Displaying the collected image of the high-voltage line in real time, and identifying obstacles on the high-voltage line based on the real-time image;

[0025] Step 3: When an obstacle is identified from the real-time image, an alarm signal is sent to the alarm to control the alarm to sound an alarm, and a stop signal is sent to the RISC-V-based terminal detection device to control the RISC-V-based terminal detection device to stop;

[0026] Step 4: Based on the acquired real-time current, the line loss value of the high-voltage line is calculated, and when the calculated line loss value is higher than the preset threshold, an alarm signal is sent to the alarm to control the alarm to sound an alarm.

[0027] In a third aspect, the present invention provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned RISC-V-based high-voltage line detection method.

[0028] The beneficial effects of the present invention are:

[0029] (1) The present invention automatically or manually controls the movement of the RISC-V-based terminal detection device on the high-voltage line to detect the real-time current at each point on the high-voltage line and collect real-time images of the high-voltage line, thereby avoiding the contact between the maintenance personnel and the high-voltage line during current collection, and to a certain extent avoiding the risk of electric shock. The detected real-time current is sent to the cloud platform through the communication module. The cloud platform performs calculations based on the real-time current, determines whether there is theft based on the calculation results, and issues an alarm when there is power theft, so that the maintenance personnel can handle it in time, further improving work efficiency.

[0030] (2) The present invention sends the collected real-time images to the terminal monitoring device, which displays them in real time and identifies obstacles based on the real-time images. When an obstacle is identified, the alarm is controlled to sound an alarm and the RISC-V-based terminal detection device is automatically controlled to stop running, so as to remind maintenance personnel to deal with the obstacle as soon as possible, which helps to improve efficiency and avoid the occurrence of dangerous situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0032] Figure 1 This is an electrical schematic diagram of a RISC-V-based high-voltage line detection system of the present invention. DETAILED DESCRIPTION

[0033] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0034] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further explanation of the present invention. Unless otherwise specified, each technical and scientific term used in this embodiment has the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0035] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0036] In the present invention, terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are relational words determined only for the convenience of describing the structural relationships of the various parts or elements of the present invention, and do not specifically refer to any part or element in the present invention and should not be understood as limitations on the present invention.

[0037] In the present invention, terms such as "fixed connection", "connected", "connection", etc. should be understood in a broad sense, indicating that it can be fixedly connected, integrally connected or detachably connected; it can be directly connected or indirectly connected through an intermediate medium. Relevant scientific research or technical personnel in this field can determine the specific meanings of the above terms in the present invention according to specific circumstances, and they should not be understood as limiting the present invention.

[0038] Embodiment 1:

[0039] like Figure 1 As shown, this embodiment provides a RISC-V-based high-voltage line detection system, including: a RISC-V-based terminal detection device, a communication module, a terminal monitoring device and a cloud platform.

[0040] (i) Communication module 3, which realizes two-way communication between terminal detection device 1 based on RISC-V and terminal monitoring device 2 and cloud platform 4, wherein communication module 3 can adopt 4G wireless communication module.

[0041] (ii) A RISC-V-based terminal detection device 1 is arranged on a high-voltage line and moves on the high-voltage line to detect a real-time current on the high-voltage line and collect a real-time image of the high-voltage line.

[0042] The RISC-V-based terminal detection device 1 adopts a split structure including an upper device and a lower device.

[0043] The upper part of the equipment includes a rotatable tire 1-1, a motor 1-2 and a RISC-V-based microcontroller module 1-3. The rotatable tire 1-1 is in contact with the high-voltage line; the RISC-V-based microcontroller module 1-3 is connected to the terminal monitoring device 2 through a communication module 3 to receive the stop and start signals issued by the terminal monitoring device 2.

[0044] The RISC-V-based microcontroller module 1-3 is connected to the tire 1-1 through the motor 1-2 to control the operation and stop of the motor 1-2 based on the received stop and start signals, thereby driving the tire 1-1 to move forward and stop on the high-voltage line.

[0045] The upper device and the lower device are connected through an automatic lock, and after the automatic lock connection, an open-close mutual inductance coil 1-4 is formed, so that the high-voltage line passes through the coil to detect the real-time current in the high-voltage line.

[0046] The RISC-V-based microcontroller module 1-3 is connected to the open-close mutual inductance coil 1-4 to obtain the detected real-time current and send it to the cloud platform 4 through the communication module 3.

[0047] Among them, a machine learning model is deployed in the RISC-V-based microcontroller module. The machine learning model is deployed in the RISC-V-based microcontroller module through model compression, model quantization and model conversion operations, which helps to reduce the model size, thereby helping to solve the problem in the prior art that large machine learning models cannot be run due to limited hardware resources of the end-side device.

[0048] It should be noted that since there are many formats of machine learning models, only machine learning models in the onnx format can be deployed on RISC-V-based microcontroller modules, so model conversion operations are required.

[0049] Specifically, the model conversion operation steps are as follows:

[0050] First, determine the model conditions of the onnx format machine learning model that can be deployed by the RISC-V-based microcontroller module, where the model conditions include data structure, storage format, and representation method of model parameters.

[0051] Secondly, model information is extracted from the unconverted machine learning model, wherein the model information includes necessary information such as architecture, weight parameters, bias terms, and activation functions.

[0052] Then, according to the model conditions and model information, the corresponding transformation algorithm is designed to complete related operations such as data reorganization, parameter encoding and decoding.

[0053] Finally, a large number of tests are performed on the converted machine learning model. Test data is input into the machine learning model before conversion and the machine learning model after conversion, and the consistency of the output results is compared to ensure the accuracy of the machine learning model format conversion.

[0054] (III) Terminal monitoring equipment:

[0055] The terminal monitoring device 2 is connected to the RISC-V-based terminal detection device 1 and the alarm 5 through the communication module 3 to obtain the image collected by the RISC-V-based terminal detection device 1 for real-time display, and identify obstacles on the high-voltage line based on the acquired image. When the existence of an obstacle on the high-voltage line is identified based on the acquired image, an alarm signal is sent to the alarm 5 to control the alarm 5 to alarm, and at the same time, a stop signal is sent to the RISC-V-based terminal detection device 1 to control the RISC-V-based terminal detection device 1 to stop.

[0056] It should be noted that the stop signal and the start signal can be sent manually or automatically to the RISC-V based terminal detection device through the terminal monitoring device 1.

[0057] Specifically, the terminal monitoring equipment includes:

[0058] The display unit 2-1 is used to display the real-time image acquired by the RISC-V-based terminal detection device 1 in real time.

[0059] The image feature extraction unit 2-2 is used to extract features from the image captured by the RISC-V-based terminal detection device 1 to generate a feature map.

[0060] The obstacle recognition unit 2-3 uses the SPP module in the YOLOv4 detection algorithm framework to identify the feature map to determine whether there is an obstacle, wherein the SPP module includes three maximum poolings of different scales and an average pooling.

[0061] Specifically, the maximum pooling of three different scales may be, for example, 1*1, 5*5, and 13*13; the average pooling may be 9*9.

[0062] In the prior art, the SPP module includes four maximum pooling operations of different scales (for example, 1*1, 5*5, 9*9, and 13*13). When the background information of the high-voltage line image is complex and similar to the obstacle, the maximum pooling operation of four different scales may only retain the background information and lose the line obstacle information during the detection process of such images, thereby causing the model to miss detection and misdetection. However, by using three maximum pooling operations and one average pooling operation, more target information of the obstacle can be retained during scale fusion, making the identification of the obstacle more accurate.

[0063] For example, when a feature map of size 13*13*1024 enters the SPP module, it no longer performs the original maximum pooling operations of size 1*1, 5*5, 9*9, 13*13, but performs three maximum pooling operations of size 1*1, 5*5, 13*13 and one average pooling operation of size 9*9. After passing through the SPP module, it enters the FPN+PAN module with 26*26*512 and 52*52*256 for feature fusion, and finally the Head module predicts the feature map, which helps to identify obstacles more accurately.

[0064] (IV) Cloud Platform 4:

[0065] The cloud platform 4 is connected to the RISC-V-based terminal detection device 1 and the alarm 5 through the communication module 3 to obtain the real-time current on the high-voltage line, and calculates the line loss value of the high-voltage line based on the obtained real-time current, and when the calculated line loss value is higher than the preset threshold, sends an alarm signal to the alarm 5 to control the alarm 5 to alarm.

[0066] The calculation of the line loss value of the high voltage line comprises the following steps:

[0067] Calculate the voltage and current relationship between the common node of this level and the common node of the next level, as well as the mutual impedance between the high-voltage lines. The specific method is as follows:

[0068] The voltage and current of three-phase electricity can be represented by vectors, where, for example, the three-phase voltage phasor of the next-level common node is recorded as Uk3.

[0069] In the plane stationary coordinate system, the vector of each phase of the three-phase voltage is:

[0070] ; ; ;

[0071] in, is the voltage amplitude, is the angular frequency.

[0072] Therefore, the three-phase voltage vector of the next-level common node can be expressed as: = + +

[0073] Based on the next level common node, the three-phase voltage vector of the common node at this level is obtained as:

[0074]

[0075] in: is the three-phase voltage phasor of the common point at this level calculated based on the common point k of the next level, is the three-phase current phasor flowing through the high-voltage line between the common node of this level and the common node of the next level; is the impedance matrix of the high voltage line between the common node of this level and the common node of the next level; The three-phase voltage phasor of the next level common node is.

[0076] It should be noted that It can be obtained based on the proportional relationship between the output voltage and the output current induced by the current transformer.

[0077] Impedance Matrix It can be solved by the injection current method or the injection voltage method. During the actual test process, the node impedance matrix can be provided by the local power supply company.

[0078] In actual use, if there is no electricity theft, the three-phase voltage "standard value U" of each common point and each load node is calculated according to different load conditions. std "; Based on the load value when the electricity theft is suspected, the voltage of each common point and the voltage of the load node are calculated: U PC+ld ; Then U PC+ld Compared with Ustd, if the voltage values ​​of several load nodes and their nearby common points are greater than a certain threshold, it is determined that there is another power theft point among these load nodes.

[0079] It should be noted that the common nodes of each branch line in the main line, the common nodes of the branch line in the branch line, and the common nodes of the branch line in the branch line in each high-voltage line are all regarded as public junctions, referred to as public nodes. The downstream of the public junction may be directly connected to the high-voltage user node, or it may be other public junctions at a lower level. In the radial (tree-shaped) distribution network, the public junction point and the non-leaf node are the same type of node, that is, non-user nodes, which means that the possibility of the public point being directly connected downstream of the public point is not ruled out.

[0080] Embodiment 2:

[0081] This embodiment provides a high-voltage line detection method based on RISC-V, comprising the following steps:

[0082] S1: Detect the real-time current on the high-voltage line and collect the real-time image of the high-voltage line.

[0083] Specifically, a start signal is manually sent to a RISC-V-based terminal detection device through a terminal monitoring device. After receiving the start signal, the RISC-V-based terminal detection device moves on the high-voltage line to detect the real-time current at each point of the high-voltage line and collect real-time images of the line.

[0084] S2: Displaying the collected real-time image, and identifying obstacles on the high-voltage line based on the real-time image.

[0085] Specifically, the RISC-V-based terminal detection device sends the collected real-time images to the terminal monitoring device through the communication module. The terminal monitoring device displays the received real-time images in real time and identifies obstacles on the high-voltage line based on the real-time images.

[0086] S3: When an obstacle is identified from the real-time image, an alarm signal is sent to the alarm to control the alarm to sound an alarm, and a stop signal is sent to the RISC-V-based terminal detection device to control the RISC-V-based terminal detection device to stop;

[0087] When the terminal monitoring device recognizes an obstacle from the real-time image, it sends an alarm signal to the alarm to control the alarm to sound an alarm, and at the same time automatically sends a stop signal to the RISC-V-based terminal detection device to control the RISC-V-based terminal detection device to stop.

[0088] S4: Based on the acquired real-time current, the line loss value of the high-voltage line is calculated, and when the calculated line loss value is higher than a preset threshold, an alarm signal is sent to the alarm to control the alarm to sound an alarm.

[0089] Specifically, the cloud platform calculates the line loss value of the new high-voltage line based on the acquired real-time current, and when the calculated line loss value is higher than a preset threshold, sends an alarm signal to the alarm to control the alarm to sound an alarm.

[0090] Embodiment 3:

[0091] This embodiment provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute a RISC-V-based high-voltage line detection method described in Example 1.

[0092] In this specification, the same or similar parts between the various embodiments can be referred to each other. In particular, for the terminal embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method embodiment.

[0093] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or units, which can be electrical, mechanical or other forms.

[0094] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0095] In addition, it should be noted that the flowchart in the accompanying drawings shows the method of the embodiment of the present disclosure. In the corresponding description in the flowchart or block diagram in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in an order different from that disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be performed substantially in parallel, and they can sometimes be performed in the opposite order, which can depend on the functions involved. Each block in the block diagram and / or flow chart, and the combination of blocks in the block diagram and / or flow chart, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A high-voltage line detection system based on RISC-V, characterized in that: include: RISC-V-based terminal detection equipment, communication modules, terminal monitoring equipment and cloud platforms; RISC-V-based terminal detection equipment is set on the high-voltage line to detect the real-time current on the high-voltage line and collect real-time images on the high-voltage line; The terminal monitoring device communicates with the RISC-V-based terminal detection device through a communication module, is used to obtain images collected by the RISC-V-based terminal detection device for real-time display, and recognize obstacles on the high-voltage line based on the acquired images; The cloud platform is connected to the RISC-V-based terminal detection equipment and alarm through the communication module to obtain the real-time current on the high-voltage line, and calculate the line loss value of the high-voltage line based on the obtained real-time current; The terminal monitoring device is also used to send a start signal and a stop signal to the RISC-V-based terminal detection device; The terminal monitoring device comprises: A display unit, used for displaying in real time the real-time image acquired by the RISC-V-based terminal detection device; An image feature extraction unit, used to extract features from an image acquired by a RISC-V-based terminal detection device to generate a feature map; The obstacle recognition unit uses the SPP module in the YOLOv4 detection algorithm framework to identify the feature map to identify obstacles. The SPP module includes three maximum poolings of different scales, namely 1*1, 5*5, and 13*13, and a 9*9 average pooling; The terminal detection device based on RISC-V includes: a real-time current detection unit; The real-time current detection unit adopts a split structure including an upper device and a lower device; The upper part of the equipment includes rotatable tires, motors and RISC-V-based microcontroller modules; A machine learning model in an onnx format is deployed in a RISC-V-based microcontroller module, wherein the machine learning model is deployed in the RISC-V-based microcontroller module through model compression, model quantization, and model conversion operations; The model conversion operation includes: First, determine the model conditions of the machine learning model in the onnx format that can be deployed by the RISC-V-based microcontroller module, wherein the model conditions include a data structure, a storage format, and a representation method of model parameters; Secondly, extracting model information from the unconverted machine learning model, wherein the model information includes necessary information of architecture, weight parameters, bias terms, and activation functions; Then, according to the model conditions and model information, the corresponding conversion algorithm is designed to complete the data reorganization, parameter encoding and decoding related operations; Finally, a large number of tests are performed on the converted machine learning model. Test data is input into the machine learning model before conversion and the machine learning model after conversion, and the consistency of the output results is compared to ensure the accuracy of the machine learning model format conversion.

2. The RISC-V-based high-voltage line detection system according to claim 1, characterized in that: The rotatable tire is in contact with the high-voltage line; the RISC-V-based microcontroller module is connected to the terminal monitoring device through the communication module to receive the stop and start signals sent by the terminal monitoring device; The RISC-V-based microcontroller module is connected to the tire through the motor to control the operation and stop of the motor based on the received stop and start signals, thereby driving the tire to move forward and stop on the high-voltage line; The upper device and the lower device are connected by an automatic lock, and after the automatic lock is connected, an open-close mutual inductance coil is formed, so that the high-voltage line passes through the coil to detect the real-time current in the high-voltage line; The RISC-V-based microcontroller module is connected to the open-and-close mutual inductance coil to obtain the detected real-time current and send it to the cloud platform through the communication module.

3. The RISC-V-based high-voltage line detection system according to claim 1, characterized in that: The communication module adopts a 4G wireless communication module.

4. A high-voltage line detection method based on RISC-V, characterized in that: The following steps are involved: Step 1: Detect the real-time current on the high-voltage line and collect the real-time image of the high-voltage line; Step 2: Displaying the collected image of the high-voltage line in real time, and identifying obstacles on the high-voltage line based on the real-time image; Step 3: When an obstacle is identified from the real-time image, an alarm signal is sent to the alarm to control the alarm to sound an alarm, and a stop signal is sent to the RISC-V-based terminal detection device to control the RISC-V-based terminal detection device to stop; Step 4: Based on the acquired real-time current, the line loss value of the high-voltage line is calculated, and when the calculated line loss value is higher than a preset threshold, an alarm signal is sent to the alarm to control the alarm to sound an alarm; The step 2 includes: displaying in real time the real-time image acquired by the RISC-V-based terminal detection device; Perform feature extraction on the image acquired by the RISC-V-based terminal detection device to generate a feature map; The SPP module in the YOLOv4 detection algorithm framework is used to identify the feature map to identify obstacles. The SPP module includes three maximum poolings of different scales, namely 1*1, 5*5, and 13*13, and a 9*9 average pooling. Before the step 1, the method further includes: deploying the machine learning model in the onnx format in a RISC-V-based microcontroller module through model compression, model quantization, and model conversion operations; The model conversion operation includes: First, determine the model conditions of the machine learning model in the onnx format that can be deployed by the RISC-V-based microcontroller module, wherein the model conditions include a data structure, a storage format, and a representation method of model parameters; Secondly, extracting model information from the unconverted machine learning model, wherein the model information includes necessary information of architecture, weight parameters, bias terms, and activation functions; Then, according to the model conditions and model information, the corresponding conversion algorithm is designed to complete the data reorganization, parameter encoding and decoding related operations; Finally, a large number of tests are performed on the converted machine learning model. Test data is input into the machine learning model before conversion and the machine learning model after conversion, and the consistency of the output results is compared to ensure the accuracy of the machine learning model format conversion.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the RISC-V-based high-voltage line detection method described in claim 4.

Images