Data detection method and system for microwave network in extra-large open-pit mines

Abstract

This application relates to the field of data processing technology, specifically to a data detection method and system for microwave networks in large open-pit mines. Through a series of carefully designed steps and strategies, it achieves comprehensive, in-depth, and accurate analysis and processing of microwave network signal data, providing strong technical support and guarantee for the optimization and maintenance of microwave networks.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a data detection method, a microwave network data detection system, and a computer-readable storage medium for microwave networks in large open-pit mines. Background Technology

[0002] Large open-pit mines, as important mineral resource extraction sites, are characterized by their vast areas and complex terrain, making traditional wired communication network cabling extremely difficult and costly to maintain. To address this issue, microwave networks, as a wireless transmission method, have been widely adopted in large open-pit mines. Microwave networks offer advantages such as long transmission distances, large bandwidth, and strong anti-interference capabilities, effectively meeting the communication needs of open-pit mining areas.

[0003] However, microwave networks can be affected by various factors during operation, such as changes in weather conditions and equipment failures, all of which can lead to a decline in network performance or even network outages. Therefore, to ensure the stable operation of microwave networks, an effective data detection technology is needed to promptly identify and resolve problems. Summary of the Invention

[0004] To address the technical problems existing in related technologies, this application provides a data detection method, a microwave network data detection system, and a computer-readable storage medium for microwave networks in large open-pit mines.

[0005] In a first aspect, embodiments of this application provide a data detection method for microwave networks in large open-pit mines, applied to a microwave network data detection system, the method comprising:

[0006] Obtain a microwave quality test report containing target microwave network signal data for analysis;

[0007] Mining the semantics of the basic microwave quality inspection elements in the microwave quality inspection report to be analyzed, the semantics of the basic microwave quality inspection elements include X basic inspection element semantic relationship networks, each basic inspection element semantic relationship network includes Y inspection element semantic units, and the inspection element semantic units with the same distributed labels in the X basic inspection element semantic relationship networks correspond to a report text paragraph of the microwave quality inspection report to be analyzed, where X and Y are positive integers;

[0008] Based on the feature-focusing strategy and the semantic relationship network of the X basic detection elements, determine the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed;

[0009] The relative distribution descriptive variables of each report text paragraph in the basic microwave quality inspection element semantics are obtained respectively, and the basic microwave quality inspection element semantics, the Y report conclusion characterization vectors and the Y relative distribution descriptive variables are accumulated into the target microwave quality inspection element semantics of the microwave quality inspection report to be analyzed.

[0010] The semantics of the target microwave quality detection elements are analyzed for signal quality inspection to obtain the signal quality inspection analysis viewpoint of the target microwave network signal data.

[0011] In conjunction with the first aspect, in one possible implementation of the first aspect, the target microwave network signal data includes microwave network command and control data, and the signal quality inspection and analysis viewpoint includes a set of command and control delay viewpoints of the microwave network command and control data;

[0012] Alternatively, the target microwave network signal data may include operation log file signal data, and the signal quality inspection and analysis viewpoints may include the instruction control delay viewpoint set of the operation log file signal data;

[0013] Alternatively, the target microwave network signal data may include signal data from a first operation log file, and the signal quality inspection and analysis viewpoint may include signal data from a second operation log file, wherein the file confidentiality tags of the first operation log file signal data and the second operation log file signal data are different.

[0014] In conjunction with the first aspect, in one possible implementation of the first aspect, determining the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed based on the feature focusing strategy and the semantic relationship network of the X basic detection elements includes:

[0015] Based on the first depth residual operator, semantic residual connections are performed on the semantic relation networks of the X basic detection elements to obtain Z residual semantic relation networks of the first detection elements, where Z is a positive integer;

[0016] Based on the second depth residual operator, semantic residual connections are performed on the semantic relationship networks of the X basic detection elements to obtain Z second detection element residual semantic relationship networks;

[0017] Based on the third deep residual operator, semantic residual connections are made on the semantic relationship networks of the X basic detection elements to obtain Z third detection element residual semantic relationship networks. The relationship network size of any first detection element residual semantic relationship network, any second detection element residual semantic relationship network, and any third detection element residual semantic relationship network is consistent with that of any basic detection element semantic relationship network.

[0018] Based on the feature-focusing strategy, the residual semantic relationship network of the Z first detection elements, the residual semantic relationship network of the Z second detection elements, and the residual semantic relationship network of the Z third detection elements are identified, and the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed are determined.

[0019] In conjunction with the first aspect, in one possible implementation of the first aspect, the selected report text paragraph is one of the Y report text paragraphs;

[0020] Based on a feature-focusing strategy, the steps for identifying the residual semantic relationship networks of the Z first detection elements, the Z second detection elements, and the Z third detection elements, and determining the report conclusion representation vector of a selected text segment in the microwave quality inspection report to be analyzed, include:

[0021] Based on the residual semantic relationship network of the Z first detection elements and the residual semantic relationship network of the Z second detection elements, determine the Y text paragraph commonality metric values ​​between the selected report text paragraph and the Y report text paragraphs;

[0022] Based on the commonality metrics of the selected report text paragraph and the Y text paragraphs, as well as the semantic relationship network of the Z third detection elements, the report conclusion representation vector of the selected report text paragraph is determined.

[0023] In conjunction with the first aspect, in one possible implementation of the first aspect, the Z second detection element residual semantic relation network includes Y second residual semantic linear knowledge, and the knowledge feature size of each second residual semantic linear knowledge is Z;

[0024] The step of determining the Y commonality metrics between the selected report text paragraph and the Y report text paragraphs based on the residual semantic relationship network of the Z first detection elements and the residual semantic relationship network of the Z second detection elements includes:

[0025] The word vector label variables of the selected report text paragraph corresponding to each word in the residual semantic relationship network of each first detection element are obtained respectively, and the obtained Z word vector label variables are integrated into the first residual semantic linear knowledge;

[0026] The first residual semantic linear knowledge is multiplied with the Y second residual semantic linear knowledge respectively to obtain the Y text paragraph commonality measure values ​​between the selected report text paragraph and the Y report text paragraphs.

[0027] In conjunction with the first aspect, in one possible implementation of the first aspect, the Z third detection element residual semantic relation network includes Y third residual semantic linear knowledge, and the knowledge feature size of each third residual semantic linear knowledge is Z;

[0028] The step of determining the report conclusion representation vector of the selected report text paragraph based on the common metrics of the Y text paragraphs and the Z third detection element residual semantic relationship network includes: performing feature enhancement processing on the common metrics of the Y text paragraphs and the Y third residual semantic linear knowledge to determine the report conclusion representation vector of the selected report text paragraph.

[0029] In conjunction with the first aspect, in one possible implementation of the first aspect, the selected report text paragraph is one of the Y report text paragraphs;

[0030] The steps for obtaining the relative distribution descriptive variables of selected report text paragraphs within the semantics of basic microwave quality inspection elements include:

[0031] Obtain the selected detection element semantic unit corresponding to the selected report text paragraph in the basic microwave quality detection element semantics;

[0032] Obtain the first selected distribution label and the second selected distribution label corresponding to the selected detection element semantic unit in the semantic relationship network of the X basic detection elements, respectively;

[0033] Based on the first selected distribution label and the second selected distribution label, a selected relative distribution descriptive variable is generated, wherein the knowledge feature size of the selected relative distribution descriptive variable is Z.

[0034] In conjunction with the first aspect, in one possible implementation of the first aspect, the semantics of the target microwave quality detection element includes Y linear semantic knowledge, and the knowledge feature size of each linear semantic knowledge is Z;

[0035] The step of performing signal quality inspection and parsing on the semantics of the target microwave quality detection elements to obtain the signal quality inspection and parsing perspective of the target microwave network signal data includes:

[0036] Based on the local feature focusing strategy and the Y linear semantic knowledge, a list of local feature focusing factors is obtained, and based on the list of local feature focusing factors, Y focused linear semantic knowledge corresponding to the Y linear semantic knowledge is determined respectively.

[0037] The viewpoint output network is invoked to parse the Y focused linear semantic knowledge respectively, and the signal quality inspection and parsing viewpoint of the target microwave network signal data is determined.

[0038] In conjunction with the first aspect, in one possible implementation of the first aspect, the method further includes:

[0039] Obtain a microwave quality inspection report sample set, which includes multiple microwave quality inspection report samples, positive signal quality inspection analysis viewpoints corresponding to each microwave quality inspection report sample, and negative signal quality inspection analysis viewpoints corresponding to each microwave quality inspection report sample;

[0040] The test report sample analysis network is invoked to obtain the first set of discriminative possibilities corresponding to each microwave quality test report sample. The test report sample analysis network includes a decision operator and an output operator.

[0041] The detection report sample analysis network is jointly perturbed and debugged based on each of the positive signal quality inspection analysis viewpoint examples, each of the negative signal quality inspection analysis viewpoint examples, and each of the first discrimination possibility sets;

[0042] When the debugged test report sample analysis network meets the network stability requirements, the debugged output operator is determined as the signal quality inspection processing network. The signal quality inspection processing network is used to identify the microwave quality test report to be analyzed that contains the target microwave network signal data, so as to obtain the signal quality inspection analysis view of the target microwave network signal data.

[0043] In conjunction with the first aspect, in one possible implementation of the first aspect, the joint perturbation and debugging of the detection report sample analysis network based on each of the positive signal quality inspection analysis viewpoint examples, each of the negative signal quality inspection analysis viewpoint examples, and each of the first discriminative probability sets includes:

[0044] Based on the positive signal quality inspection analysis viewpoint example of each microwave quality inspection report example, the first discriminative possibility set of each microwave quality inspection report example, and the negative signal quality inspection analysis viewpoint example of each microwave quality inspection report example, the algorithm weight of the decision operator is improved;

[0045] Based on the improved decision operator, obtain the second discriminant probability set corresponding to each microwave quality inspection report sample;

[0046] Based on the positive signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample and the second discriminative possibility set of each microwave quality inspection report sample, the algorithm weight of the output operator is improved, and the decision operator and the output operator are cyclically debugged.

[0047] Furthermore, in conjunction with the first aspect, in an independent implementation of the first aspect, the improvement of the algorithm weights of the decision operator based on the positive signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample, the first discriminative probability set of each microwave quality inspection report sample, and the negative signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample includes:

[0048] Based on the positive signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample and the first discriminant probability set of each microwave quality inspection report sample, the first network training error is determined;

[0049] The second network training error is determined based on the first set of discriminative possibilities for each microwave quality inspection report sample.

[0050] Based on the negative signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample and the first discriminant probability set of each microwave quality inspection report sample, the training error of the third network is determined;

[0051] The first network training error, the second network training error, and the third network training error are accumulated into the target network training error, and the algorithm weights of the decision operator are improved in reverse order based on the target network training error.

[0052] Furthermore, in conjunction with the first aspect, in an independent implementation of the first aspect, the improvement of the algorithm weights of the output operator based on the positive signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample and the second discriminative probability set of each microwave quality inspection report sample includes:

[0053] Based on the positive signal quality inspection analysis viewpoint example of each microwave quality inspection report example, and the second discriminant possibility set of each microwave quality inspection report example, the target expected offset variable is obtained;

[0054] The algorithm weights of the output operator are improved by reversing the target expected offset variable.

[0055] Secondly, this application also provides a microwave network data detection system, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include methods for executing any of the data detection methods for microwave networks in large open-pit mines.

[0056] Thirdly, this application also provides a computer-readable storage medium, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute any of the data detection methods for microwave networks in large open-pit mines.

[0057] This application proposes a data detection method based on semantic analysis and feature focusing for microwave networks in large open-pit mines. First, a microwave quality inspection report containing target microwave network signal data is obtained, providing a data foundation for subsequent analysis. Next, the semantics of the basic microwave quality inspection elements in the report are mined. These basic inspection element semantics constitute an X-fold semantic relationship network, with each network consisting of Y inspection element semantic units. This structured representation allows the system to more clearly understand and interpret the report content. To further improve the accuracy and efficiency of the analysis, Y report conclusion representation vectors are determined for the Y report text paragraphs in the microwave quality inspection report based on a feature focusing strategy and the X-fold semantic relationship network. These vectors capture the key information and features of the report, facilitating subsequent signal quality inspection analysis. Then, the relative distribution descriptive variables of each report text paragraph in the basic microwave quality inspection element semantics are obtained. These variables reflect the relative positions and relationships of different elements in the overall distribution, providing an important basis for comprehensively evaluating network performance. Finally, the semantics of the basic microwave quality inspection elements, Y report conclusion representation vectors, and Y relative distribution descriptive variables are accumulated to form the semantics of the target microwave quality inspection elements in the microwave quality inspection report to be analyzed. Ultimately, signal quality inspection analysis is performed on the semantics of the target microwave quality inspection elements to obtain signal quality inspection analysis perspectives for the target microwave network signal data. These perspectives can provide targeted guidance for network optimization and maintenance, ensuring the stable operation of the microwave network in large open-pit mines. Attached Figure Description

[0058] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0059] Figure 1 A hardware structure block diagram of a mobile terminal for performing a data detection method for a microwave network in a large open-pit mine, according to an embodiment of this application, is shown.

[0060] Figure 2 A schematic flowchart of a data detection method for a microwave network in a large open-pit mine, according to an embodiment of this application, is shown.

[0061] The above figures include the following reference numerals:

[0062] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0063] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0064] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0065] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0066] As described in the background section, microwave networks in the prior art may be affected by various factors during operation, such as changes in weather conditions and equipment failures. These factors may lead to a decline in network performance or even an interruption. To solve the above problems, embodiments of this application provide a data detection method, a microwave network data detection system, and a computer-readable storage medium for microwave networks in large open-pit mines.

[0067] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0068] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a data detection method of a microwave network in a large open-pit mine, according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0069] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the device information display method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0070] This embodiment provides a data detection method for a microwave network in a large open-pit mine that runs on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0071] Based on this, please refer to Figure 2 , Figure 2This is a flowchart illustrating a data detection method for a microwave network in a large open-pit mine, provided in an embodiment of this application. The method is applied to a microwave network data detection system and may further include steps 210-250.

[0072] Step 210: The microwave network data detection system obtains a microwave quality detection report containing the target microwave network signal data to be analyzed.

[0073] Step 220: The microwave network data detection system mines the semantic elements of the basic microwave quality detection elements in the microwave quality detection report to be analyzed.

[0074] Among them, the semantics of basic microwave quality inspection elements include X basic inspection element semantic relationship networks, and any basic inspection element semantic relationship network includes Y inspection element semantic units. Inspection element semantic units with the same distribution labels in the X basic inspection element semantic relationship networks correspond to a report text segment of the microwave quality inspection report to be analyzed above, where X and Y are positive integers.

[0075] Step 230: The microwave network data detection system determines the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality detection report to be analyzed based on the feature focusing strategy and the semantic relationship network of the above X basic detection elements.

[0076] Step 240: The microwave network data detection system obtains the relative distribution description variables of each report text paragraph in the semantics of the basic microwave quality detection elements, and accumulates the above-mentioned semantics of the basic microwave quality detection elements, the above-mentioned Y report conclusion representation vectors and Y relative distribution description variables into the semantics of the target microwave quality detection elements of the microwave quality detection report to be analyzed.

[0077] Step 250: The microwave network data detection system performs signal quality inspection and analysis on the semantics of the above-mentioned target microwave quality detection elements to obtain the signal quality inspection and analysis viewpoints of the above-mentioned target microwave network signal data.

[0078] Some microwave network data inspection systems are automated, efficient data analysis tools specifically designed to evaluate and improve the quality and performance of microwave networks. The following is a specific application scenario demonstrating how this system operates according to a predefined technical solution.

[0079] The microwave network data detection system first retrieves a microwave quality detection report containing the target microwave network signal data from a central database or real-time data stream. This report details various signal parameters in the network, such as signal strength, noise level, and transmission delay.

[0080] The system uses natural language processing and machine learning algorithms to mine the semantics of the basic microwave quality detection elements in this report. These semantics are organized into X basic detection element semantic relationship networks, each network containing Y detection element semantic units. These units and the relationship networks together constitute the textual structure and information content of the report. For example, a relationship network might represent the detection element "signal strength," while its contained semantic units might be specific signal strength values, trends in change, or relationships with other parameters.

[0081] Next, based on the feature-focusing strategy and the previously established semantic relationship network of basic detection elements, the system determines Y report conclusion representation vectors for Y report text paragraphs in the report. These vectors are highly abstract data structures that can capture the core information and features of each text paragraph, facilitating subsequent analysis.

[0082] The system continues its work, acquiring the relative distribution descriptive variables of each report text paragraph within the semantics of the basic microwave quality inspection elements. These variables reflect the inherent connections and differences between different text paragraphs. Subsequently, the system accumulates the semantics of the basic microwave quality inspection elements, Y report conclusion representation vectors, and Y relative distribution descriptive variables to form the semantics of the target microwave quality inspection elements in the microwave quality inspection report to be analyzed. This is a comprehensive and detailed dataset, covering all the key information and features of the report.

[0083] Finally, the system performs signal quality inspection analysis on the semantics of the target microwave quality detection elements. This process involves complex algorithms and models, enabling in-depth analysis of the inherent patterns and trends in the data, thereby deriving signal quality inspection analysis insights for the target microwave network signal data. These insights are highly specialized and instructive conclusions that can provide strong support for the optimization and management of microwave networks. For example, the system may identify problems such as insufficient signal strength, excessively high noise levels, or excessively long transmission delays in certain areas, and propose corresponding improvement measures and optimization suggestions.

[0084] Furthermore, the following is a more detailed description of an implementation example of a microwave network data detection system.

[0085] Detailed Explanation of Microwave Network Data Detection System Implementation:

[0086] In complex environments such as large open-pit mines, microwave networks serve as the backbone of data transmission, and their stability and performance are crucial. Microwave network data monitoring systems were developed for this purpose, enabling real-time monitoring and analysis of the microwave network's status and providing decision support for network optimization.

[0087] The system first automatically captures a comprehensive microwave quality inspection report from the central database or real-time data stream. This report details various signal parameters of the microwave network within a specific time period, including but not limited to key indicators such as signal strength, signal-to-noise ratio, bit error rate, and transmission delay. This raw data forms the basis for subsequent analysis.

[0088] To fully understand the report's content, the system employs natural language processing and deep learning techniques for in-depth analysis. It identifies key terms and phrases in the report, such as "signal attenuation" and "noise interference," and categorizes these terms into different semantic relationship networks of basic detection elements. Each relationship network represents a specific detection dimension, such as signal quality or network stability. The semantic units within each relationship network further refine the specific indicators and states of these dimensions.

[0089] After understanding the basic semantics of the report, the system performs high-level abstraction on each paragraph of the report text according to a preset feature-focusing strategy. It extracts the core information of each paragraph, such as anomaly indicators and trends, and encodes this information into representation vectors. These vectors not only capture the semantic features of the text but also preserve the logical relationships and contextual information between paragraphs.

[0090] Next, the system enters the integration phase. It first calculates the relative distribution descriptive variables of each report text paragraph within the semantics of the basic microwave quality inspection elements. These variables reflect the degree to which different paragraphs contribute to the overall semantics. Then, the system organically combines the basic semantics, the representation vector, and the relative distribution descriptive variables to form a comprehensive and detailed semantic model of the target microwave quality inspection elements. This model not only includes the report's raw data and information but also incorporates the system's understanding and analysis results.

[0091] Finally, the system performs in-depth signal quality inspection analysis on the semantic model of the target microwave quality inspection elements. It employs complex algorithms and models to uncover inherent patterns and potential problems in the data, such as signal interference patterns and performance bottlenecks. Based on these analytical results, the system generates professional and instructive signal quality inspection analysis viewpoints. These viewpoints not only point out the current problems and challenges in microwave networks but also provide targeted optimization suggestions and improvement measures. For example, the system might suggest adjusting the configuration parameters of certain devices, optimizing the network topology, or adding signal relay stations.

[0092] Through the above process, the microwave network data detection system can achieve comprehensive monitoring and in-depth analysis of microwave networks. It not only improves network operation and maintenance efficiency and quality but also provides strong support for intelligent and safe production in mines. In the future, with continuous technological advancements and increasing application demands, this system is expected to play an even greater role in more fields.

[0093] In the embodiments of this application, the exemplary explanations of the technical terms appearing in the above technical solutions are as follows.

[0094] Target microwave network signal data refers to the signal data that needs to be monitored and analyzed within a specific microwave network. This data typically includes key parameters such as signal strength, frequency, phase, noise level, and bit error rate, and is crucial for evaluating the performance and quality of the microwave network. For example, in a microwave communication network in an open-pit mine, target microwave network signal data might include signal strength readings between various transmission nodes, bit error rate statistics, and signal delay measurements. This data is essential for ensuring the reliability and real-time performance of communication within the mine.

[0095] Microwave Quality Inspection Report to be Analyzed: This report contains the signal quality inspection results of the microwave network. It is typically generated by specialized testing equipment or systems and details the performance and quality status of the microwave network under specific time or conditions. For example, a microwave quality inspection report to be analyzed might include a signal strength distribution map of the mine's microwave communication network over the past week, a bit error rate trend chart, and a network stability analysis report. This report will serve as input to the microwave network data inspection system for further data analysis and problem diagnosis.

[0096] Basic microwave quality inspection element semantics: Basic microwave quality inspection element semantics refers to the meaning and explanation of the fundamental concepts, terms, and parameters involved in microwave quality inspection reports. These semantic elements are the foundation for understanding the report content and analyzing microwave network performance. For example, in a microwave quality inspection report, "signal strength" is a basic microwave quality inspection element semantic, representing the power or intensity of the microwave signal. Other common basic microwave quality inspection element semantics include "noise level," "bit error rate," and "transmission delay."

[0097] The semantic relationship network of basic detection elements is a network structure composed of multiple related semantic elements of basic microwave quality detection. These elements are connected by some logical relationship or mutual influence relationship, jointly describing a certain aspect or level of microwave network performance. For example, in microwave network performance evaluation, "signal strength" and "transmission distance" can constitute a semantic relationship network of basic detection elements. Because signal strength usually attenuates with the increase of transmission distance, there is a close correlation and influence between these two elements.

[0098] Semantic Unit of Detection Element: A semantic unit of detection element is the smallest unit in the basic semantic relationship network of detection elements, representing a specific detection element or parameter. Each semantic unit of detection element contains basic information about that element, its attributes, and its relationships with other elements. For example, in the basic semantic relationship network of detection elements composed of "signal strength" and "transmission distance," "signal strength" and "transmission distance" are themselves two semantic units of detection elements. They each have different attribute values ​​and relationship descriptions, together forming the basic framework of this relationship network.

[0099] Distribution Labels: Distribution labels are used to identify and classify the distribution characteristics of microwave network signal data across specific dimensions. These labels help analysts quickly identify and understand data distribution patterns, outliers, and potential performance issues. For example, in microwave network signal data, "signal strength distribution" can serve as a distribution label. Based on this label, analysts can categorize signal strength data into three levels: "strong," "medium," and "weak," and then implement different optimization measures or processing strategies for each level.

[0100] Report text paragraphs: These paragraphs are part of the microwave quality inspection report being analyzed. They typically consist of a set of related sentences or statements used to describe and analyze the performance and quality status of the microwave network in a specific aspect or dimension. For example, in a microwave quality inspection report being analyzed, the "Signal Strength Analysis" section may contain one or more report text paragraphs. These paragraphs detail the signal strength measurement methods, results, and possible influencing factors, providing analysts with comprehensive signal strength information and data support.

[0101] Feature-focusing strategy: This strategy focuses on extracting the most relevant and representative features for the target task when processing and analyzing data. In microwave network data inspection, feature-focusing strategies can help the system more effectively understand and interpret report content, thereby improving the accuracy and efficiency of the analysis. For example, when analyzing the performance of a microwave network, a feature-focusing strategy might emphasize extracting signal parameters that directly affect network performance, such as signal strength, noise level, and transmission delay, while ignoring information less relevant to the current analysis task, such as equipment brand or model.

[0102] Report Conclusion Representation Vector: A report conclusion representation vector refers to the process of converting the conclusions or key information of a report into a mathematical vector form using a specific algorithm or model. These vectors capture the semantic features and key information of the report, facilitating subsequent mathematical analysis and calculations. For example, in a microwave quality inspection report, the conclusion section might indicate that the signal quality in a certain area is poor. Through natural language processing and machine learning techniques, the system can convert this conclusion into a representation vector, which may contain information about signal quality, area location, and other relevant factors.

[0103] Relative distribution descriptive variables: These variables characterize the relative position, proportion, or relationship of a particular element or feature within the overall distribution. In microwave network data detection, these variables help analysts understand the relative importance and interrelationships between different signal parameters. For example, when analyzing the signal intensity distribution of a microwave network, relative distribution descriptive variables can include the mean, standard deviation, maximum, minimum, and proportions of different intensity ranges. These variables comprehensively describe the distribution characteristics and patterns of signal intensity.

[0104] Semantics of Target Microwave Quality Inspection Elements: Target microwave quality inspection element semantics refers to the semantic meaning of the elements or characteristics that need to be focused on and analyzed in a specific microwave quality inspection task. These element semantics are the foundation for understanding report content, evaluating network performance, and developing optimization strategies. For example, when evaluating the stability of a microwave network, the semantics of target microwave quality inspection elements might include the frequency and amplitude of signal fluctuations, the number of failures, and their duration. Accurate understanding and analysis of these element semantics are crucial for assessing network stability and reliability.

[0105] Signal quality inspection and analysis: This refers to the process of conducting in-depth quality inspection and analysis of microwave network signal data using specialized techniques and algorithms. This process aims to identify problems, anomalies, or potential risks in the signal, providing decision support for network optimization and maintenance. For example, in a microwave network data inspection system, signal quality inspection and analysis may include multiple aspects such as signal strength analysis, noise level assessment, bit error rate calculation, and transmission delay measurement. Through these analyses, the system can generate detailed quality inspection reports, pointing out problems in the network and providing improvement suggestions.

[0106] Signal quality inspection and analysis perspectives: These perspectives are conclusions or judgments drawn from the results of signal quality inspection and analysis. These perspectives directly reflect and evaluate the signal quality status of the microwave network, and are of great significance for guiding network optimization and maintenance. For example, when analyzing signal data of a microwave network, signal quality inspection and analysis perspectives may indicate problems such as insufficient signal strength, severe noise interference, or excessive transmission delay in a certain area. To address these issues, the system can further propose optimization suggestions, such as adjusting equipment layout, adding signal repeaters, or improving the transmission environment.

[0107] As can be seen, by applying the above embodiments, firstly, step 210 obtains a microwave quality detection report containing the target microwave network signal data, providing a comprehensive and accurate data foundation for subsequent in-depth analysis and processing. This step ensures that the starting point of the analysis is reliable and targeted, thereby improving the accuracy and effectiveness of the overall analysis.

[0108] Secondly, step 220 mines the semantics of the basic microwave quality inspection elements in the microwave quality inspection report to be analyzed and constructs a complex model including a semantic relationship network of X basic inspection elements. Each basic inspection element semantic relationship network is composed of Y inspection element semantic units. This meticulous division enables the system to more accurately understand and interpret the report content, providing a solid foundation for subsequent signal quality assessment and optimization.

[0109] Third, step 230, based on a feature-focusing strategy and the semantic relationship network of X basic detection elements, determines Y report conclusion representation vectors for Y report text paragraphs in the microwave quality inspection report to be analyzed. This step effectively transforms textual information into a computable mathematical form by focusing on key features, facilitating subsequent quantitative analysis. Simultaneously, this significantly improves the efficiency and accuracy of the analysis.

[0110] Fourth, step 240 obtains the relative distribution descriptive variables of each report text paragraph in the semantics of the basic microwave quality inspection elements, and accumulates the semantics of the basic microwave quality inspection elements, the Y report conclusion representation vectors, and the Y relative distribution descriptive variables into the semantics of the target microwave quality inspection elements of the microwave quality inspection report to be analyzed. This step achieves the organic integration of multi-dimensional information, enabling the system to understand the report content more comprehensively and deeply, providing strong support for signal quality inspection analysis.

[0111] Finally, step 250 performs signal quality analysis on the semantics of the target microwave quality detection elements to obtain the signal quality analysis viewpoint of the target microwave network signal data. This step further refines and elevates the previous analysis results, providing guiding decision support for network optimization and maintenance. This also demonstrates the value and significance of this application in practical applications.

[0112] In summary, this application, through a series of carefully designed steps and strategies, achieves comprehensive, in-depth, and accurate analysis and processing of microwave network signal data, providing strong technical support and assurance for the optimization and maintenance of microwave networks.

[0113] In some examples, the target microwave network signal data includes microwave network command and control data, and the signal quality inspection and analysis viewpoint includes a set of command and control delay viewpoints for the microwave network command and control data; or, the target microwave network signal data includes operation log file signal data, and the signal quality inspection and analysis viewpoint includes a set of command and control delay viewpoints for the operation log file signal data; or, the target microwave network signal data includes first operation log file signal data, and the signal quality inspection and analysis viewpoint includes second operation log file signal data, wherein the file confidentiality tags of the first operation log file signal data and the second operation log file signal data are different.

[0114] In some exemplary embodiments, the target microwave network signal data processed by this system can have various different types, each corresponding to a specific signal quality inspection and analysis perspective. The following will provide detailed examples of these different types of target microwave network signal data and their corresponding signal quality inspection and analysis perspectives.

[0115] First, when the target microwave network signal data includes microwave network command and control data, the signal quality inspection and analysis perspective of this system will focus on the command and control latency of this data. Specifically, the system will construct a set of perspectives on command and control latency by deeply mining and analyzing the transmission characteristics and quality of the microwave network command and control data. This set of perspectives will describe in detail the transmission latency of command and control data in the network, including but not limited to information such as latency time, frequency, trend, and possible causes. In this way, the system can effectively monitor and evaluate the transmission performance of microwave network command and control data, providing strong data support for network optimization and maintenance.

[0116] Secondly, when the target microwave network signal data includes signal data from the operation log files, the system's signal quality inspection and analysis perspective will also focus on the command and control latency of these log files. Operation log files typically record various states and events during microwave network operation, serving as a crucial window into network performance. The system will carefully analyze the signal data in these operation log files, extracting key information related to command and control latency and constructing a corresponding command and control latency perspective set. This perspective set will describe in detail the command and control latency events recorded in the operation log files, including the time of occurrence, cause, scope of impact, and solutions. In this way, the system can help network administrators promptly identify and resolve command and control latency issues in the network, improving network stability and reliability.

[0117] Finally, when the target microwave network signal data includes signal data from the first operation log file, the system's signal quality inspection and analysis perspective involves signal data from the second operation log file, and the confidentiality tags for these two types of operation log file signal data are different. In this case, the system first performs in-depth analysis and processing on the signal data from the first and second operation log files separately, extracting their respective key information and features. Then, the system combines the different confidentiality tags of these two files to construct a signal quality inspection and analysis perspective regarding file confidentiality. This perspective details the differences and connections between the two operation log files in terms of confidentiality, and the potential impact of these differences on network performance and security. In this way, the system can provide network administrators with a more comprehensive and in-depth perspective on network performance and security analysis, helping them better understand and manage microwave networks.

[0118] In summary, by flexibly processing different types of target microwave network signal data and constructing corresponding signal quality inspection and analysis viewpoints, this system can provide comprehensive and accurate data support and analytical perspectives for the optimization and maintenance of microwave networks. Whether it's the transmission delay of microwave network command and control data, or the differences in command and control delay events and file confidentiality in the operation log file signal data, this system can perform in-depth analysis and processing, providing strong support for the stable operation and continuous improvement of the network.

[0119] In the next step, the feature-focusing strategy described in step 230 and the semantic relationship network of the X basic detection elements are used to determine the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed, including steps 231-234.

[0120] Step 231: Perform semantic residual connection on the semantic relationship network of the above X basic detection elements according to the first depth residual operator to obtain Z first detection element residual semantic relationship networks, where Z is a positive integer.

[0121] Step 232: Perform semantic residual connection on the semantic relationship network of the above X basic detection elements according to the second depth residual operator to obtain the semantic relationship network of the Z second detection elements residual.

[0122] Step 233: Perform semantic residual connections on the semantic relationship networks of the above X basic detection elements according to the third depth residual operator to obtain Z third detection element residual semantic relationship networks. The relationship network scale of any first detection element residual semantic relationship network, any second detection element residual semantic relationship network, and any third detection element residual semantic relationship network is consistent with that of any basic detection element semantic relationship network.

[0123] Step 234: Based on the feature focusing strategy, identify the residual semantic relationship network of the Z first detection elements, the residual semantic relationship network of the Z second detection elements, and the residual semantic relationship network of the Z third detection elements, and determine the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed.

[0124] In the following steps, step 230 is further refined into steps 231 to 234, in order to describe in more detail how to determine the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality inspection report to be analyzed based on the feature focusing strategy and the semantic relationship network of X basic detection elements.

[0125] In step 231, the system performs semantic residual connections on the semantic relationship network of X basic detection elements based on the first deep residual operator. The deep residual operator is a commonly used technique in deep learning, which helps the network better learn the complex relationships between input and output during training. Through the action of the first deep residual operator, the system obtains Z residual semantic relationship networks of the first detection elements. These residual semantic relationship networks, while preserving the original semantic information, further highlight certain key features, which is helpful for subsequent analysis and processing.

[0126] Step 232 is similar to step 231, but uses the second deep residual operator. By performing semantic residual connections on the semantic relationship networks of X basic detection elements, the system obtains Z second detection element residual semantic relationship networks. The second deep residual operator may have different characteristics and concerns than the first deep residual operator, therefore the resulting second detection element residual semantic relationship networks will also differ. This difference helps the system understand and interpret the report content from multiple perspectives and levels.

[0127] In step 233, the system again uses the deep residual operator, this time the third deep residual operator, to perform semantic residual connections on the semantic relation network of the X basic detection elements. Through this process, the system obtains Z residual semantic relation networks of the third detection elements. At this point, the system has obtained three different sets of residual semantic relation networks, each containing Z relation networks. These relation networks are consistent in scale with the original semantic relation network of the basic detection elements, but differ in semantic expression and emphasis.

[0128] In step 234, the system identifies and processes the three sets of residual semantic relationship networks obtained earlier based on a feature-focusing strategy. The feature-focusing strategy is a targeted analysis method that emphasizes the identification and utilization of key features. By applying the feature-focusing strategy, the system can identify the features and information most relevant to the report conclusions from the residual semantic relationship networks of Z first detection elements, Z second detection elements, and Z third detection elements. Then, based on these features and information, the system determines Y report conclusion representation vectors for Y report text paragraphs in the microwave quality inspection report to be analyzed. These representation vectors mathematically express the conclusive viewpoints and information of the report text paragraphs, providing important input data for subsequent signal quality inspection analysis.

[0129] Furthermore, the beneficial effects of steps 231 to 234 are mainly reflected in the following aspects:

[0130] Feature enhancement and highlighting: Through the action of the deep residual operator, steps 231, 232, and 233 generate the first, second, and third residual semantic relationship networks of the detected elements, respectively. These residual semantic relationship networks, while preserving the original semantic information, further enhance the representation of key features, making these features more prominent in subsequent analysis and helping to improve the accuracy and efficiency of the analysis;

[0131] Multi-perspective analysis: By using different deep residual operators, information in the report can be captured from multiple angles and levels. Each deep residual operator may focus on different features or relationships, thus generating three sets of residual semantic relationship networks that provide diverse interpretations of the report, helping the system to understand the report content more comprehensively;

[0132] Improved robustness: The introduction of deep residual operators helps alleviate the gradient vanishing or exploding problems during the training process of deep learning models, thereby improving the model's robustness. This enhanced robustness helps ensure the stability and reliability of the analysis when processing complex microwave quality inspection reports.

[0133] Feature Focusing and Dimensionality Reduction: Step 234 uses a feature focusing strategy to identify the features most relevant to the reported conclusions from a large number of features, achieving feature focusing and dimensionality reduction. This helps reduce the complexity and computational cost of subsequent analysis, while ensuring that the analysis focuses on the most critical features;

[0134] Improving parsing accuracy: Through the preceding steps, the system can more accurately capture key information and features in the report, thereby generating more precise report conclusion representation vectors in step 234. These representation vectors, as input for subsequent signal quality inspection parsing, will help improve the accuracy and effectiveness of the analysis.

[0135] In summary, steps 231 to 234, by introducing a deep residual operator and a feature focusing strategy, enable multi-level and multi-angle analysis of microwave quality inspection reports, improving the accuracy, efficiency, and robustness of the analysis, and providing strong support for subsequent signal quality inspection and analysis.

[0136] In some examples, the selected report text paragraph is one of the aforementioned Y report text paragraphs. Then, step 234, which describes a feature-focused strategy to identify the residual semantic relationship networks of the Z first detection elements, the Z second detection elements, and the Z third detection elements, and to determine the report conclusion representation vector of the selected report text paragraph in the microwave quality inspection report to be analyzed, includes steps 2341-2342.

[0137] Step 2341: Based on the residual semantic relationship network of the Z first detection elements and the residual semantic relationship network of the Z second detection elements, determine the Y commonality measure values ​​of the selected report text paragraphs and the Y report text paragraphs.

[0138] Step 2342: Based on the commonality measure values ​​of the selected report text paragraph and the Y text paragraphs, as well as the semantic relationship network of the residuals of the Z third detection elements mentioned above, determine the report conclusion representation vector of the selected report text paragraph.

[0139] In some examples, when the system selects a specific report text paragraph as the analysis object, step 234 will be further refined into steps 2341 and 2342 to determine the report conclusion representation vector of the selected report text paragraph.

[0140] In step 2341, the system first utilizes the previously generated residual semantic relationship networks of Z first detection elements and Z second detection elements to determine the commonality measure between the selected report text paragraph and Y other report text paragraphs. This commonality measure can be understood as the degree of semantic and content similarity or consistency between the selected report text paragraph and other report text paragraphs. By calculating these commonality measures, the system can identify the correlation and consistency between the selected report text paragraph and other text paragraphs in the entire report, providing an important reference for subsequently determining the report conclusion representation vector.

[0141] Specifically, the system may employ a similarity measurement algorithm (such as cosine similarity, Euclidean distance, etc.) to quantitatively evaluate the semantic similarity between the selected report text paragraph and each other report text paragraph. These evaluation results will constitute Y commonality metrics for text paragraphs, reflecting the commonalities and differences in content between the selected report text paragraph and the entire report.

[0142] In step 2342, the system will determine the report conclusion representation vector of the selected report text paragraph based on the Y commonality metrics of the text paragraphs calculated in step 2341 and the previously generated Z residual semantic relationship networks of the third detection elements. This representation vector will be a multi-dimensional mathematical expression used to describe the key features of the selected report text paragraph in terms of content, semantics, and conclusion.

[0143] Specifically, the system may employ a weighted fusion strategy to combine common metrics with information from the semantic relationship network of the residuals of the third-detection elements. For example, the system can adjust the weights of different features in the semantic relationship network of the residuals of the third-detection elements based on the magnitude of the common metrics, so that features more similar to and consistent with the entire report are more prominently represented in the representation vector. Through this processing method, the system can generate a report conclusion representation vector that contains both unique information from the selected report text paragraph and reflects its relevance to the entire report.

[0144] In another example, the aforementioned Z second-detection element residual semantic relationship network includes Y second-residual semantic linear knowledge, each of which has a knowledge feature size of Z. Step 2341, which describes determining the Y commonality metrics between the selected report text paragraph and the Y report text paragraphs based on the aforementioned Z first-detection element residual semantic relationship network and the aforementioned Z second-detection element residual semantic relationship network, includes: obtaining the word vector label variables corresponding to the selected report text paragraphs in each first-detection element residual semantic relationship network, and integrating the obtained Z word vector label variables into first-residual semantic linear knowledge; performing feature multiplication processing on the aforementioned first-residual semantic linear knowledge with the aforementioned Y second-residual semantic linear knowledge to obtain the Y commonality metrics between the selected report text paragraph and the Y report text paragraphs.

[0145] In another example, when the Z second detection element residual semantic relationship network includes Y second residual semantic linear knowledge, and the knowledge feature size of each second residual semantic linear knowledge is Z, the implementation of step 2341 will use a specific method to determine the commonality metric between the selected report text paragraph and the Y report text paragraphs.

[0146] First, the system accesses the residual semantic relationship network of each first detection element and extracts the word vector label variables of the corresponding words in the selected report text paragraph. These word vector label variables are the representations of each word in the text paragraph in the semantic space, capturing the semantic relationships and contextual information between words. By extracting these variables, the system can obtain the expression of the selected report text paragraph at different semantic levels.

[0147] Next, the system integrates these Z word vector label variables into a first residual semantic linear knowledge. This integration process may involve weighted averaging, concatenation, or other forms of combination of the word vector label variables to form a linear knowledge representation that can comprehensively reflect the semantic features of the selected report text paragraphs.

[0148] Then, the system performs feature multiplication on the first residual semantic linear knowledge and Y second residual semantic linear knowledge respectively. Feature multiplication is a feature fusion technique that captures the interactions and nonlinear relationships between different features by calculating the product of different features. In this step, the system uses feature multiplication to combine the first residual semantic linear knowledge with each second residual semantic linear knowledge to explore the commonalities and correlations between the selected report text paragraph and each other report text paragraph.

[0149] Finally, through feature multiplication, the system obtained Y commonality metrics between the selected report text paragraph and Y other report text paragraphs. These metrics quantify the degree of semantic similarity and content consistency between the selected report text paragraph and the other report text paragraphs. These commonality metrics will provide an important basis for subsequently determining the report conclusion representation vector of the selected report text paragraph.

[0150] In another example, the aforementioned Z third-factor residual semantic relationship network includes Y third-factor residual semantic linear knowledge, each of which has a knowledge feature size of Z. Therefore, step 2342, which describes determining the report conclusion representation vector of the selected report text paragraph based on the Y text paragraph commonality metrics between the selected report text paragraph and the aforementioned Z third-factor residual semantic relationship network, includes: performing feature enhancement processing on the aforementioned Y text paragraph commonality metrics and the aforementioned Y third-factor residual semantic linear knowledge to determine the report conclusion representation vector of the selected report text paragraph.

[0151] In another example, when the Z third detection element residual semantic relationship network contains Y third residual semantic linear knowledge, and the knowledge feature size of each third residual semantic linear knowledge is Z, step 2342 will use feature enhancement processing to determine the report conclusion representation vector of the selected report text paragraph.

[0152] Feature enhancement is a technique that increases the influence of specific features in the final representation. In this step, the system first considers the commonality metrics of the Y text paragraphs calculated earlier. These commonality metrics reflect the degree of similarity and consistency between the selected report text paragraphs and other report text paragraphs, and therefore have important reference value in determining the report conclusion representation vector.

[0153] Next, the system will combine these common metrics with Y third residual semantic linear knowledge points. Specifically, the system may employ weighted averaging, element-wise multiplication, or other fusion strategies, using the common metrics as weights or adjustment factors applied to the corresponding third residual semantic linear knowledge points. This approach ensures that the third residual semantic linear knowledge points that are more similar to and consistent with the selected report text paragraphs are more fully represented in the final representation.

[0154] Through feature enhancement processing, the system can effectively combine the unique information of selected report text paragraphs with the common features of the entire report, generating a report conclusion representation vector that contains both individual characteristics and reflects commonalities. This representation vector will serve as a high-level abstraction of the semantics and content of the selected report text paragraphs, providing strong support for subsequent signal quality inspection and analysis.

[0155] In some possible embodiments, the selected report text paragraph is one of the Y report text paragraphs; then the step of obtaining the relative distribution descriptive variable of the selected report text paragraph in the semantics of the basic microwave quality inspection elements includes: obtaining the selected inspection element semantic unit corresponding to the selected report text paragraph in the semantics of the basic microwave quality inspection elements; obtaining the first selected distribution label and the second selected distribution label corresponding to the selected inspection element semantic unit in the semantic relationship network of the X basic inspection elements respectively; generating a selected relative distribution descriptive variable based on the first selected distribution label and the second selected distribution label, wherein the knowledge feature size of the selected relative distribution descriptive variable is Z.

[0156] In some possible embodiments, when the system needs to obtain the relative distribution description variables of selected report text paragraphs in the semantics of basic microwave quality inspection elements, it performs this task in a series of steps.

[0157] First, the system locates the semantic unit of the selected report text paragraph within the basic microwave quality inspection element semantics. This process is similar to finding a specific term or concept in a large text library; however, here, the system searches for the semantic unit corresponding to the selected report text paragraph within a predefined set of microwave quality inspection element semantics.

[0158] Next, the system will further obtain the first and second selected distribution labels of this selected detection element semantic unit in the semantic relationship network of X basic detection elements. Here, the "semantic relationship network of basic detection elements" can be understood as a pre-constructed knowledge graph or network structure that describes the complex relationships and interdependencies between microwave quality detection elements. The "first selected distribution label" and "second selected distribution label" may refer to the different positions or roles of this semantic unit in this network structure, or its specific association with other semantic units.

[0159] Specifically, the first selected distribution label may represent the global position or importance of the selected detection element semantic unit within the semantic relationship network of the basic detection elements, while the second selected distribution label may focus more on describing its local relationships or interactions with other specific semantic units. These labels are typically obtained through complex algorithms and extensive data analysis, and they help the system more accurately understand and interpret the meaning and context of selected report text paragraphs.

[0160] Finally, based on these first and second selected distribution labels, the system generates a selected relative distribution descriptive variable. This variable is actually a multi-dimensional mathematical expression that integrates various features and attributes of the selected detection element semantic unit within the semantics of the basic microwave quality detection element. Its knowledge feature size is Z, meaning that this variable contains Z different dimensions or features, each reflecting information or attributes of the selected report text paragraph in a specific aspect or level.

[0161] This selected relative distribution descriptive variable is one of the key results of the system's in-depth understanding and analysis of the selected report text paragraph. It not only helps the system more accurately grasp the theme and key points of the paragraph, but also provides strong support and reference for subsequent signal quality inspection analysis. In this way, the system can achieve comprehensive, accurate, and efficient processing and analysis of microwave quality inspection reports.

[0162] In some other embodiments, the semantics of the target microwave quality detection elements mentioned above include Y linear semantic knowledge, and the knowledge feature size of each linear semantic knowledge is Z; then the signal quality inspection and analysis of the semantics of the target microwave quality detection elements described in step 250 to obtain the signal quality inspection and analysis viewpoint of the target microwave network signal data includes steps 251-252.

[0163] Step 251: Based on the local feature focusing strategy and the above Y linear semantic knowledge, obtain the list of local feature focusing factors and, based on the list of local feature focusing factors, determine the Y focused linear semantic knowledge corresponding to the Y linear semantic knowledge respectively.

[0164] Step 252: Call the viewpoint output network to parse the above Y focused linear semantic knowledge respectively, and determine the signal quality inspection and parsing viewpoint of the above target microwave network signal data.

[0165] In other embodiments, when the system needs to perform signal quality inspection parsing on the semantics of target microwave quality inspection elements, it executes a series of detailed steps to obtain signal quality inspection parsing insights. These steps include steps 251 and 252, which together constitute an in-depth analysis and understanding of the semantics of the target microwave quality inspection elements.

[0166] First, step 251 requires the system to obtain a list of local feature focusing factors based on a local feature focusing strategy and Y linear semantic knowledge points. The local feature focusing strategy is a method that emphasizes the importance of features within a specific region, helping the system to more accurately identify and process key information. In this step, the system analyzes each linear semantic knowledge point in the semantics of the target microwave quality detection elements and generates a list of local feature focusing factors based on their importance and relevance. This list contains factors used to emphasize or highlight specific features, which will guide subsequent processing and analysis.

[0167] Next, the system uses this list of local feature focusing factors to determine the Y focused linear semantic knowledge corresponding to each of the Y linear semantic knowledge pieces. Focusing is a technique that concentrates attention on the most important or relevant features, helping to improve the system's sensitivity and processing capabilities for key information. By applying the list of local feature focusing factors, the system can highlight the key features in each linear semantic knowledge piece and generate the corresponding focused linear semantic knowledge. This focused linear semantic knowledge will serve as the basis for subsequent signal quality inspection and analysis.

[0168] Then, in step 252, the system invokes the viewpoint output network to parse these Y focused linear semantic knowledge points. The viewpoint output network is a neural network structure specifically designed to parse and extract specific viewpoints or opinions. It can learn from large amounts of data to identify and understand different semantic patterns and viewpoint expressions. In this step, the viewpoint output network receives the focused linear semantic knowledge as input and uses its complex internal structure and algorithms to parse this input, extracting the signal quality checks and parsing the viewpoints contained within.

[0169] Ultimately, through the parsing and processing of the viewpoint output network, the system determines the signal quality inspection viewpoints for the target microwave network signal data. These viewpoints are the result of a deep understanding and analysis of the semantics of the target microwave quality inspection elements, reflecting the system's perception and judgment of the target microwave network signal data quality and performance. These signal quality inspection viewpoints will provide important references and basis for subsequent decision-making and optimization.

[0170] In some other alternative embodiments, the method further includes steps 310-340.

[0171] Step 310: Obtain a set of microwave quality inspection report samples. The set of microwave quality inspection report samples includes multiple microwave quality inspection report samples, positive signal quality inspection analysis viewpoint samples corresponding to each microwave quality inspection report sample, and negative signal quality inspection analysis viewpoint samples corresponding to each microwave quality inspection report sample.

[0172] Step 320: Call the test report sample analysis network to obtain the first discrimination probability set corresponding to each microwave quality test report sample. The test report sample analysis network includes decision operators and output operators.

[0173] Step 330: Perform joint perturbation debugging on the above detection report sample analysis network based on each of the above positive signal quality inspection analysis viewpoint examples, each of the above negative signal quality inspection analysis viewpoint examples, and each of the above first discrimination probability sets.

[0174] Step 340: When the debugged test report sample analysis network meets the network stability requirements, the debugged output operator is determined as the signal quality inspection processing network. The above-mentioned signal quality inspection processing network is used to identify the microwave quality test report to be analyzed that contains the target microwave network signal data, so as to obtain the signal quality inspection analysis view of the above-mentioned target microwave network signal data.

[0175] In some alternative embodiments, the system further extends its functionality by including steps 310 to 340, which together constitute a complete process for improving the accuracy of microwave quality inspection report analysis.

[0176] First, in step 310, the system acquires a set of microwave quality inspection report samples. This set not only contains multiple microwave quality inspection report samples, but also provides corresponding positive signal quality inspection analysis viewpoints and negative signal quality inspection analysis viewpoints for each sample. These samples provide the system with a foundation for learning and reference, helping the system understand how to extract positive and negative signal quality inspection analysis viewpoints from different microwave quality inspection reports.

[0177] Next, in step 320, the system invokes a neural network model called the "Detection Report Sample Analysis Network." This network model comprises two main components: a decision operator and an output operator. The decision operator is responsible for performing preliminary analysis and judgment based on the input microwave quality inspection report samples, while the output operator is responsible for transforming the results of these analyses and judgments into specific signal quality inspection analytical viewpoints. In this step, the system uses this network model to obtain the first set of discriminative possibilities corresponding to each microwave quality inspection report sample. This set reflects the system's preliminary judgment on whether each sample may belong to a positive or negative category.

[0178] Then, in step 330, the system performs joint perturbation debugging. This process combines each positive signal quality inspection analysis viewpoint example, each negative signal quality inspection analysis viewpoint example, and each first discriminant probability set to meticulously adjust and optimize the detection report sample analysis network. By continuously adjusting the parameters and structure of the network model, the system can gradually improve its analysis accuracy of microwave quality inspection reports, making its output signal quality inspection analysis viewpoints closer to the actual positive or negative categories.

[0179] Finally, in step 340, the system evaluates the stability of the debugged test report sample analysis network. Only when this network model demonstrates stable performance in multiple consecutive tests and meets the preset network stability requirements will the system determine its output operator as the signal quality inspection processing network. This signal quality inspection processing network is the core tool used by the system to ultimately identify the microwave quality test report to be analyzed, which contains target microwave network signal data. Through it, the system can quickly extract signal quality inspection analysis points from the target microwave network signal data, providing strong support for subsequent decision-making and optimization.

[0180] In the following steps, the joint perturbation and debugging of the detection report sample analysis network based on each of the above-mentioned positive signal quality inspection analysis viewpoints, each of the above-mentioned negative signal quality inspection analysis viewpoints, and each of the above-mentioned first discriminant probability sets, as described in step 330, includes: improving the algorithm weights of the decision operator based on the above-mentioned positive signal quality inspection analysis viewpoints, the above-mentioned first discriminant probability sets, and the above-mentioned negative signal quality inspection analysis viewpoints of each microwave quality detection report sample; obtaining the second discriminant probability set corresponding to each microwave quality detection report sample based on the improved decision operator; and improving the algorithm weights of the output operator based on the above-mentioned positive signal quality inspection analysis viewpoints and the second discriminant probability sets of each microwave quality detection report sample. The decision operator and the output operator are cyclically debugged.

[0181] In the following steps, the joint perturbation debugging process described in step 330 is a complex and delicate process involving iterative improvement of the algorithm weights of the decision and output operators in the detection report sample analysis network. This process is based on the positive signal quality inspection analytical viewpoint sample, the first discriminative probability set, and the negative signal quality inspection analytical viewpoint sample for each microwave quality inspection report sample.

[0182] First, the system improves the algorithm weights of the decision operator based on the positive signal quality inspection analysis viewpoints, the first set of discriminative possibilities, and the negative signal quality inspection analysis viewpoints for each microwave quality inspection report sample. Specifically, the system analyzes the differences and consistency among these samples, and adjusts the weight allocation of the decision operator when processing similar reports by comparing the correspondence between positive and negative viewpoint samples and the first set of discriminative possibilities. In this way, the decision operator can make a more accurate preliminary judgment based on the characteristics of the input microwave quality inspection report sample.

[0183] Then, the system uses the improved decision operator to re-analyze each microwave quality inspection report sample, obtaining a second set of discriminative possibilities for each sample. This set reflects the new judgment result of the decision operator after weight adjustment, indicating whether each sample may belong to the positive or negative category.

[0184] Next, the system further improves the algorithm weights of the output operator based on the positive signal quality inspection analysis viewpoint examples and the second set of discriminative possibilities for each microwave quality inspection report example. Similar to the improvement of the decision operator, the system adjusts the weight allocation of the output operator when generating the final signal quality inspection analysis viewpoint by analyzing the correspondence between the positive viewpoint examples and the second set of discriminative possibilities. In this way, the output operator can more accurately generate a signal quality inspection analysis viewpoint that conforms to the actual situation based on the judgment results of the decision operator.

[0185] It is important to note that the debugging of the decision operator and the output operator is a cyclical process. The system will continuously repeat the above steps until the preset debugging termination conditions are met (such as reaching the maximum number of iterations or meeting certain performance improvement requirements). During this process, the algorithm weights of the decision operator and the output operator will be gradually optimized, continuously improving the accuracy of the detection report sample analysis network in analyzing microwave quality inspection reports. Ultimately, the fully debugged detection report sample analysis network will be able to more accurately identify the microwave quality inspection report to be analyzed that contains target microwave network signal data and output corresponding signal quality inspection analysis opinions.

[0186] In the following steps, the algorithm weights of the decision operator are improved based on the positive signal quality inspection analysis viewpoints of each microwave quality inspection report sample, the first discriminant probability set of each microwave quality inspection report sample, and the negative signal quality inspection analysis viewpoints of each microwave quality inspection report sample. This includes: determining a first network training error based on the positive signal quality inspection analysis viewpoints of each microwave quality inspection report sample and the first discriminant probability set of each microwave quality inspection report sample; determining a second network training error based on the first discriminant probability set of each microwave quality inspection report sample; determining a third network training error based on the negative signal quality inspection analysis viewpoints of each microwave quality inspection report sample and the first discriminant probability set of each microwave quality inspection report sample; accumulating the first network training error, the second network training error, and the third network training error into a target network training error; and improving the algorithm weights of the decision operator in reverse order based on the target network training error.

[0187] In the next step, the system improves the algorithm weights of the decision operator based on the positive signal quality inspection analysis viewpoint sample, the first set of discriminative possibilities, and the negative signal quality inspection analysis viewpoint sample for each microwave quality inspection report sample. This process is achieved by calculating different types of network training errors and accumulating them into the target network training error.

[0188] First, the system determines the first network training error based on the positive signal quality inspection parsing viewpoint sample and the first set of discriminative possibilities for each microwave quality inspection report sample. This error reflects the accuracy of the decision operator in judging the positive signal quality inspection parsing viewpoint. Specifically, the system compares the difference between the positive signal quality inspection parsing viewpoint sample and the first set of discriminative possibilities to calculate the error value of the decision operator when judging it as a positive category.

[0189] Secondly, the system determines the second network training error based on the first set of discriminant possibilities for each microwave quality inspection report sample. This error reflects the accuracy of the decision operator in judging the overall category of the microwave quality inspection report sample. The system calculates the error value of the decision operator in the overall judgment based on the difference between the first set of discriminant possibilities and the actual category label.

[0190] Next, the system determines the third network training error based on the negative signal quality inspection analysis viewpoint sample and the first set of discriminative possibilities for each microwave quality inspection report sample. This error reflects the accuracy of the decision operator in judging the negative signal quality inspection analysis viewpoint. Similar to calculating the first network training error, the system compares the difference between the negative signal quality inspection analysis viewpoint sample and the first set of discriminative possibilities to calculate the error value of the decision operator when judging it as a negative category.

[0191] Finally, the system accumulates the training errors of the first, second, and third networks into a target network training error. This target network training error comprehensively reflects the overall accuracy of the decision operator in judging microwave quality inspection report samples. Then, the system reverse-orders the algorithm weights of the decision operator based on the target network training error. Reverse-order improvement means that the system prioritizes adjusting those weights that cause larger errors, in order to gradually reduce the target network training error and improve the judgment accuracy of the decision operator.

[0192] Through this process, the algorithm weights of the decision operator are optimized, enabling the test report sample analysis network to more accurately determine the category of microwave quality test reports in subsequent processing and output corresponding signal quality analysis opinions.

[0193] In the following steps, the algorithm weights of the output operator are improved based on the positive signal quality inspection analysis viewpoint sample and the second discriminant probability set of each microwave quality inspection report sample, including: obtaining the target expected offset variable based on the positive signal quality inspection analysis viewpoint sample and the second discriminant probability set of each microwave quality inspection report sample; and improving the algorithm weights of the output operator in reverse order based on the target expected offset variable.

[0194] In the next step, the system improves the algorithm weights of the output operator based on the positive signal quality inspection analysis viewpoint sample and the second discriminant probability set of each microwave quality inspection report sample. This process is mainly achieved by obtaining the target expected offset variable and improving the algorithm weights of the output operator in reverse order based on this variable.

[0195] First, the system uses the positive signal quality inspection analytical viewpoint sample and the second discrimination possibility set for each microwave quality inspection report sample to obtain the target expected offset variable. This variable reflects the deviation between the judgment result of the current output operator and the positive signal quality inspection analytical viewpoint sample. Specifically, the system analyzes the correspondence between the positive signal quality inspection analytical viewpoint sample and the second discrimination possibility set, and calculates the difference between the expected output and the actual output when the output operator is judged as positive, i.e., the target expected offset variable.

[0196] After obtaining the target expected offset variable, the system will improve the algorithm weights of the output operator in reverse order based on this variable. Reverse order improvement means that the system will prioritize adjusting weights that cause larger deviations, gradually reducing the target expected offset variable and improving the judgment accuracy of the output operator. Specifically, the system will adjust the algorithm weights of the output operator according to the magnitude and direction of the target expected offset variable. If the deviation of a certain weight is large, the system will give a larger adjustment; if the deviation of a certain weight is small, the system will give a smaller adjustment or remain unchanged.

[0197] Through this process, the algorithm weights of the output operators are optimized, enabling the detection report sample analysis network to more accurately generate corresponding signal quality inspection analysis viewpoints based on the characteristics of the microwave quality inspection report in subsequent processing. In this way, the system can more accurately identify the microwave quality inspection report to be analyzed that contains target microwave network signal data, and output signal quality inspection analysis viewpoints that conform to the actual situation, providing strong support for subsequent decision-making and optimization.

[0198] This solution also provides a microwave network data detection system, including: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include methods for executing any of the above-described data detection methods for microwave networks in large open-pit mines.

[0199] Furthermore, a computer storage medium is also provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute any of the above-described data detection methods for microwave networks in large open-pit mines.

[0200] This invention provides a processor for running a program, wherein the program executes the data detection method for microwave networks in large open-pit mines.

[0201] This invention provides a device that includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements at least the steps of a data detection method for a microwave network in a large open-pit mine.

[0202] The devices mentioned in this article can be servers, PCs, tablets, mobile phones, etc.

[0203] A computer program product includes a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the data detection method for microwave networks in large open-pit mines described in various embodiments of this application.

[0204] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0205] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0206] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0207] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0208] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0209] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0210] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0211] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0212] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0213] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A data detection method for microwave networks in large open-pit mines, characterized in that, The method, applied to a microwave network data detection system, includes: Obtain a microwave quality test report containing target microwave network signal data for analysis; Mining the semantics of the basic microwave quality inspection elements in the microwave quality inspection report to be analyzed, the semantics of the basic microwave quality inspection elements include X basic inspection element semantic relationship networks, each basic inspection element semantic relationship network includes Y inspection element semantic units, and the inspection element semantic units with the same distributed labels in the X basic inspection element semantic relationship networks correspond to a report text paragraph of the microwave quality inspection report to be analyzed, where X and Y are positive integers; Based on the feature-focusing strategy and the semantic relationship network of the X basic detection elements, Y report conclusion representation vectors are determined for the Y report text paragraphs in the microwave quality inspection report to be analyzed. Specifically, the semantic relationship network of the X basic detection elements is semantically residually connected using a first deep residual operator to obtain Z residual semantic relationship networks of the first detection elements, where Z is a positive integer; the semantic relationship network of the X basic detection elements is semantically residually connected using a second deep residual operator to obtain Z residual semantic relationship networks of the second detection elements; and the semantic relationship network of the X basic detection elements is semantically residually connected using a third deep residual operator. Semantic residual connections are performed to obtain Z residual semantic relationship networks of third detection elements. The relationship network size of any residual semantic relationship network of first detection elements, any residual semantic relationship network of second detection elements, and any residual semantic relationship network of third detection elements is consistent with that of any basic detection element semantic relationship network. Based on the feature focusing strategy, the residual semantic relationship networks of the Z first detection elements, the Z residual semantic relationship networks of second detection elements, and the Z residual semantic relationship networks of third detection elements are identified to determine the Y report conclusion representation vectors of the Y report text paragraphs in the microwave quality detection report to be analyzed. The relative distribution descriptive variables of each report text paragraph in the basic microwave quality inspection element semantics are obtained respectively, and the basic microwave quality inspection element semantics, the Y report conclusion characterization vectors and the Y relative distribution descriptive variables are accumulated into the target microwave quality inspection element semantics of the microwave quality inspection report to be analyzed. The semantics of the target microwave quality detection elements are analyzed for signal quality inspection to obtain the signal quality inspection analysis viewpoint of the target microwave network signal data.

2. The method according to claim 1, characterized in that, The target microwave network signal data includes microwave network command and control data, and the signal quality inspection and analysis viewpoints include the command and control delay viewpoint set of the microwave network command and control data. Alternatively, the target microwave network signal data may include operation log file signal data, and the signal quality inspection and analysis viewpoints may include the instruction control delay viewpoint set of the operation log file signal data; Alternatively, the target microwave network signal data may include signal data from a first operation log file, and the signal quality inspection and analysis viewpoint may include signal data from a second operation log file, wherein the file confidentiality tags of the first operation log file signal data and the second operation log file signal data are different.

3. The method according to claim 1, characterized in that, The selected report text paragraph is one of the Y report text paragraphs; Based on a feature-focusing strategy, the steps for identifying the residual semantic relationship networks of the Z first detection elements, the Z second detection elements, and the Z third detection elements, and determining the report conclusion representation vector of a selected report text segment in the microwave quality inspection report to be analyzed, include: Based on the residual semantic relationship network of the Z first detection elements and the residual semantic relationship network of the Z second detection elements, determine the Y text paragraph commonality metric values ​​between the selected report text paragraph and the Y report text paragraphs; Based on the commonality metrics of the selected report text paragraph and the Y text paragraphs, as well as the semantic relationship network of the Z third detection elements, the report conclusion representation vector of the selected report text paragraph is determined.

4. The method according to claim 3, characterized in that, The Z second-detection element residual semantic relationship network includes Y second-residual semantic linear knowledge, and the knowledge feature size of each second-residual semantic linear knowledge is Z; The step of determining the Y commonality metrics between the selected report text paragraph and the Y report text paragraphs based on the residual semantic relationship network of the Z first detection elements and the residual semantic relationship network of the Z second detection elements includes: The word vector label variables of the corresponding word units of the selected report text paragraph in the residual semantic relationship network of each first detection element are obtained respectively, and the obtained Z word vector label variables are integrated into the first residual semantic linear knowledge; The first residual semantic linear knowledge is multiplied with the Y second residual semantic linear knowledge respectively to obtain the Y text paragraph commonality measure values ​​between the selected report text paragraph and the Y report text paragraphs; The Z third-factor residual semantic relationship network includes Y third-factor residual semantic linear knowledge, and the knowledge feature size of each third-factor residual semantic linear knowledge is Z. The step of determining the report conclusion representation vector of the selected report text paragraph based on the Y text paragraph commonality metric values ​​between the selected report text paragraph and the Y third-factor residual semantic relationship network includes: performing feature enhancement processing on the Y text paragraph commonality metric values ​​and the Y third-factor residual semantic linear knowledge to determine the report conclusion representation vector of the selected report text paragraph.

5. The method according to claim 1, characterized in that, The selected report text paragraph is one of the Y report text paragraphs; The steps for obtaining the relative distribution descriptive variables of selected report text paragraphs within the semantics of basic microwave quality inspection elements include: Obtain the selected detection element semantic unit corresponding to the selected report text paragraph in the basic microwave quality detection element semantics; Obtain the first selected distribution label and the second selected distribution label corresponding to the selected detection element semantic unit in the semantic relationship network of the X basic detection elements, respectively; Based on the first selected distribution label and the second selected distribution label, a selected relative distribution descriptive variable is generated, wherein the knowledge feature size of the selected relative distribution descriptive variable is Z.

6. The method according to claim 1, characterized in that, The target microwave quality detection element semantics includes Y linear semantic knowledge, and the knowledge feature size of each linear semantic knowledge is Z; The step of performing signal quality inspection and parsing on the semantics of the target microwave quality detection elements to obtain the signal quality inspection and parsing perspective of the target microwave network signal data includes: Based on the local feature focusing strategy and the Y linear semantic knowledge, a list of local feature focusing factors is obtained, and based on the list of local feature focusing factors, Y focused linear semantic knowledge corresponding to the Y linear semantic knowledge is determined respectively. The viewpoint output network is invoked to parse the Y focused linear semantic knowledge respectively, and the signal quality inspection and parsing viewpoint of the target microwave network signal data is determined.

7. The method according to claim 1, characterized in that, The method further includes: Obtain a microwave quality inspection report sample set, which includes multiple microwave quality inspection report samples, positive signal quality inspection analysis viewpoints corresponding to each microwave quality inspection report sample, and negative signal quality inspection analysis viewpoints corresponding to each microwave quality inspection report sample; The test report sample analysis network is invoked to obtain the first set of discriminative possibilities corresponding to each microwave quality test report sample. The test report sample analysis network includes a decision operator and an output operator. The detection report sample analysis network is jointly perturbed and debugged based on each of the positive signal quality inspection analysis viewpoint examples, each of the negative signal quality inspection analysis viewpoint examples, and each of the first discrimination possibility sets; When the debugged test report sample analysis network meets the network stability requirements, the debugged output operator is determined as the signal quality inspection processing network. The signal quality inspection processing network is used to identify the microwave quality test report to be analyzed that contains the target microwave network signal data, so as to obtain the signal quality inspection analysis view of the target microwave network signal data. The step of jointly perturbing and adjusting the detection report sample analysis network based on each of the positive signal quality inspection analysis viewpoint examples, each of the negative signal quality inspection analysis viewpoint examples, and each of the first discriminative probability sets includes: Based on the positive signal quality inspection analysis viewpoint example of each microwave quality inspection report example, the first discriminative possibility set of each microwave quality inspection report example, and the negative signal quality inspection analysis viewpoint example of each microwave quality inspection report example, the algorithm weight of the decision operator is improved; Based on the improved decision operator, obtain the second discriminant probability set corresponding to each microwave quality inspection report sample; Based on the positive signal quality inspection analysis viewpoint sample of each microwave quality inspection report sample and the second discriminative possibility set of each microwave quality inspection report sample, the algorithm weight of the output operator is improved, and the decision operator and the output operator are cyclically debugged.

8. A microwave network data detection system, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being used to perform the data detection method for a microwave network in a large open-pit mine as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the data detection method for microwave networks in large open-pit mines as described in any one of claims 1 to 7.

Images