Screen status fault query and processing method, system and terminal device

Abstract

The present invention relates to the field of screen maintenance and management technology, in particular to a screen status fault query and processing method, system and terminal device. In the dynamic timing feature field, the screen status word vector changes of the first screen status program sequence and the second screen status program sequence after dynamic timing update are analyzed to obtain the first screen status; by obtaining the first future screen status timing model diagram and the second future screen status timing model diagram output in the target dynamic timing feature field when the main playback device is in the future playback period, the screen status output in the target dynamic timing feature field is analyzed to obtain whether it is synchronized, and the second screen status is obtained. The present invention can quickly query and accurately resolve screen status faults between playback devices, thereby avoiding the phenomenon of templates without prices and calling numbers without sound between devices, and improving the synchronization and stability of the screen status of playback devices.

Description

Technical Field

[0001] The present invention relates to the technical field of screen maintenance management, and in particular to a screen status fault query and processing method, system and terminal equipment. Background Art

[0002] In certain specific application scenarios, it is usually necessary to use a movable or fixed sub-playback device to connect to the main playback device through the network to synchronize the playback content and interact in real time, such as using a mobile phone to connect to the restaurant's order preview terminal, so as to achieve the effect of joint screen ordering and price preview playback, which is convenient for users' lifestyles. However, if there are some faults between the playback devices that affect the joint screen playback, it will cause the screen status between the playback devices to have playback timing anomalies and delays, and traditional fault query methods are difficult to accurately query and locate the cause of the abnormal screen status, so it is difficult to formulate targeted maintenance and repair plans for processing, and the fault query and traceability performance is poor, which reduces the stability and quality of the screen linkage playback between the playback devices. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a screen status fault query and processing method, system and terminal equipment.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A first aspect of the present invention provides a screen status fault query and processing method, comprising the following steps:

[0006] S102: Obtain a first screen state program sequence of a main playback device and a second screen state program sequence of a sub-playback device, analyze the screen state word vector changes of the first screen state program sequence and the second screen state program sequence after dynamic time sequence update in the dynamic time sequence feature field, and obtain a first screen state;

[0007] S104: acquiring a first future screen state timing model diagram and a second future screen state timing model diagram output in a target dynamic timing feature field when the main playback device is in a future playback period, so as to analyze whether the screen states output in the target dynamic timing feature field are synchronized, and obtain a second screen state;

[0008] S106: obtaining several types of joint screen non-playing faults that cause the first screen state and the second screen state to be generated between the main playback device and the sub-playback device, introducing the naive Bayes theorem to calculate and analyze the several types of joint screen non-playing faults, and deploying tests using natural language processing technology to obtain screen state fault query results;

[0009] S108: If the screen status fault query result output by the mobile application terminal is displayed as the fault query result of the first screen status, then after repairing the first screen status, the current request response data is obtained, and the difference analysis between the current request response data and the historical request response data is calculated to edit and change the modification calling code to update the first screen status request code of the partner API interface list;

[0010] S110: If the screen status fault query result output by the mobile application terminal is displayed as the fault query result of the second screen status, unqualified assessment items are planned by constructing a network connection evaluation model and the normal status evaluation radar chart area corresponding to each evaluation item, and the target connection abnormality factors that cause the second screen status are determined based on the unqualified evaluation items and repaired using the best maintenance plan.

[0011] More specifically, the step S102 includes the following steps:

[0012] Obtain a main playback device and one or more sub-playback devices interconnected with the main playback device, and identify a screen state program sequence of the main playback device, defined as a first screen state program sequence, and a screen state program sequence of the sub-playback device, defined as a second screen state program sequence, through a mobile application terminal recognition technology;

[0013] Acquire a text reading format of the mobile application terminal, and perform text conversion on the first screen state program sequence and the second screen state program sequence based on the text reading format to obtain first screen state text data and second screen state text data; wherein the mobile application terminal includes enterprise WeChat and a Wed end webpage;

[0014] Introducing a stem feature extraction algorithm, performing word segmentation preprocessing on the first screen state text data and the second screen state text data, and performing stem feature extraction on the first screen state text data and the second screen state text data after the word segmentation preprocessing in the stem feature extraction algorithm to obtain a plurality of first screen state word vectors and a plurality of second screen state word vectors;

[0015] Constructing a dynamic time series feature domain, presetting N mapping nodes in the dynamic time series feature domain based on the plurality of first screen state word vectors and the plurality of second screen state word vectors, and planning a feature mapping sub-domain according to positions of the N mapping nodes in the dynamic time series feature domain;

[0016] Obtaining a dynamic time series interval corresponding to the feature mapping sub-field through a dynamic time series feature field, recording a screen state word vector after dynamic time series update on each mapping node based on the dynamic time series interval, and obtaining an updated first screen state word vector and an updated second screen state word vector;

[0017] The cosine similarity method is introduced to calculate the degree of fit between the updated first screen state word vector and the updated second screen state word vector. If the degree of fit is less than the preset degree of fit, it indicates that there is a phenomenon that the template price is not displayed on a joint screen between the main playback device and the sub-playback device.

[0018] The screen linkage state between the main playback device and the sub-playback device is marked as a type of linked screen non-playback abnormal state and uploaded to the mobile application terminal to generate a first screen state.

[0019] More specifically, the step S104 includes the following steps:

[0020] When there is no first screen status fault query result between the main playback device and the sub-playback device, then the content playback plan and the current playback timestamp of the main playback device are obtained;

[0021] Preset a future playback period, take the current playback timestamp as the starting time node, obtain multiple sets of first screen state text data of the main playback device in the future playback period based on the content playback plan and perform time sequence integration, and define it as a first future screen state text data set;

[0022] An ARIMA model algorithm is introduced to calculate the autoregressive order, the number of differences and the moving average order of the first future screen state text data set, and the autoregressive integrated moving average model is fitted through the autoregressive order, the number of differences and the moving average order to obtain the first future screen state time series dynamic model;

[0023] Setting a target dynamic timing node interval based on the current playback timestamp and the future playback timestamp, acquiring a target dynamic timing feature field corresponding to the target dynamic timing node region in the dynamic timing feature field, and dividing the target dynamic timing feature field into a plurality of uniform sub-fields;

[0024] Acquire the dynamic timing specifications of each sub-field in the dynamic timing feature field, embed the first future screen state timing dynamic model into the target dynamic timing feature field, uniformly time slice the first future screen state timing dynamic model according to the dynamic timing specifications of each sub-field, and obtain the first future screen state timing model diagram output by the main playback device in each sub-field;

[0025] Based on the above steps, the second future screen state timing model diagram output by the sub-playback device in each sub-field is calculated and obtained, and the first future screen state timing model diagram and the second future screen state timing model diagram corresponding to each sub-field are aligned. After the alignment is completed, the alignment coincidence degree between the first future screen state timing model diagram and the second future screen state timing model diagram in each sub-field is calculated;

[0026] Only sub-fields with a registration overlap less than a preset registration overlap are marked, which are defined as screen status abnormal sub-fields. The number of screen status abnormal sub-fields is counted. If the number of screen status abnormal sub-fields is greater than the preset number of sub-fields, it indicates that there is a phenomenon of no sound when the main playback device and the sub-playback device are connected to the screen;

[0027] The screen linkage state between the main playback device and the sub-playback device is marked as a second-class linked screen non-playing state and uploaded to the mobile application terminal to generate a second screen state.

[0028] More specifically, the step S106 includes the following steps:

[0029] Obtain the first historical screen state word vector corresponding to the first screen state between the main playback device and the sub-playback device, which is defined as the first historical word vector training set, and obtain the second historical screen state word vector corresponding to the second screen state between the main playback device and the sub-playback device, which is defined as the second historical word vector training set;

[0030] Acquire a network communication protocol between a main playback device and a sub-playback device, and retrieve and acquire, based on the network communication protocol, several types of joint screen non-playback failures that cause a first screen state and a second screen state to be generated between the main playback device and the sub-playback device in a big data network;

[0031] Count the occurrence frequencies of each type of joint screen non-playback failure in the first historical word vector training set and the second historical word vector training set, respectively, to obtain a plurality of first occurrence frequencies and second occurrence frequencies;

[0032] Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the first historical word vector training set according to the first occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the first screen state, which are defined as the first prior probability and the first conditional probability distribution;

[0033] Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the second historical word vector training set according to the second occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the second screen state, which are defined as the second prior probability and the second conditional probability distribution;

[0034] The naive Bayes theorem is introduced to calculate the first prior probability and the first conditional probability distribution, the second prior probability and the second conditional probability distribution, respectively, to obtain the first posterior probability and the second posterior probability, and only the joint screen non-playing fault type corresponding to the maximum first posterior probability and the maximum second posterior probability is extracted to generate the fault location query result;

[0035] A natural language processing model framework for the mobile application terminal is constructed through natural language processing technology, and the fault location query result is imported into the natural language processing model framework for deployment testing to obtain the recall rate. If the recall rate is lower than the preset recall rate, the fault location query result is re-predicted and adjusted until the recall rate reaches the preset recall rate to generate a screen status fault query result.

[0036] More specifically, the step S108 includes the following steps:

[0037] If the screen state fault query result output by the mobile application terminal is displayed as a fault query result of the first screen state, then editing and generating fault avoidance prompt information according to the first screen state word vector and the second screen state word vector;

[0038] After the user repairs the first screen state between the main playback device and the sub-playback device according to the fault avoidance prompt information, reacquire the repaired first screen state word vector and the repaired second screen state word vector and calculate the degree of fit to obtain the repaired fit;

[0039] If the repaired fit is still less than the preset fit, construct an API request text parameter according to the repaired fit, obtain the current partner API interface list, send a request to the current partner API interface list based on the API request text parameter, and generate current request response data;

[0040] Preset a pre-repair historical update node, obtain historical request response data at the pre-repair historical update node, calculate and analyze the mean of the request path field between the current request response data and the historical request response data by variance analysis method, and obtain a mean square error value;

[0041] If the mean square error value is greater than the preset mean square error value, the deviation between the mean square error value and the preset mean square error value is calculated to obtain the mean square error deviation value, and at the same time, a list of partner API interfaces at the historical update node before the repair is obtained, which is defined as a historical partner API interface list;

[0042] Analyze the interface field differences between the current partner API interface list and the historical partner API interface list according to the mean square deviation value to obtain the partner API change content, and determine the calling code interval that needs to be modified in the current partner API interface list according to the partner API change content;

[0043] Edit the modified calling code based on the partner API change content, and apply the modified calling code to the calling code interval that needs to be modified to generate a new API calling document, update the request code of the current partner API interface list through the new API calling document, and generate updated request content;

[0044] The updated request content is uploaded to the mobile application terminal to automatically query and confirm and update the service request of the playback device in real time according to the change status of the partner API interface.

[0045] More specifically, the step S110 includes the following steps:

[0046] If the screen state fault query result output by the mobile application terminal is displayed as the fault query result of the second screen state, a current network connection node between the main playback device and the sub-playback device and a preset network communication channel on the current network connection node are obtained;

[0047] Based on the current network connection node, the preset network communication channel and the network communication protocol, a plurality of evaluation items for evaluating the network connection environment between the main playback device and the sub-playback device are obtained in the big data network, and a network connection evaluation model is created according to the plurality of evaluation items;

[0048] Build a radar chart architecture, obtain the lowest evaluation threshold of each evaluation item in a normal state based on the big data network, and plan the normal state evaluation radar chart area corresponding to each evaluation item in the radar chart architecture according to the lowest evaluation threshold;

[0049] A global evaluation is performed on the network connection environment between the current main playback device and the sub-playback device through a network connection evaluation model to obtain multiple evaluation indexes, and an actual evaluation radar chart is drawn in a radar chart framework based on the multiple evaluation indexes, and the negotiation area value between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area is calculated;

[0050] If the negotiation area value is smaller than the preset negotiation area value, the normal state assessment radar chart area is marked as an unqualified assessment item, and the assessment index corresponding to the unqualified assessment item is obtained;

[0051] Based on the big data network, all the connection abnormal factors that cause the network connection to produce unqualified evaluation items are obtained, and the fuzzy C-means algorithm is introduced. Based on the evaluation index, the clusters corresponding to each unqualified evaluation item are preset in the fuzzy C-means algorithm, and the membership matrix of each connection abnormal factor for each cluster is initialized;

[0052] Preset an iterative change threshold, calculate the center point of each cluster based on the membership matrix, obtain a number of cluster centers, continuously update the iterative membership matrix through the cluster centers, and obtain the membership matrix change after the update iteration. If the membership matrix change after the update iteration is less than the iterative change threshold, stop the iterative update, and finally output the target connection abnormality factor that causes the unqualified assessment item;

[0053] Obtain several processing solutions for target connection abnormality factors and the repair rate of each processing solution, extract the processing solution corresponding to the maximum repair rate and upload it to the mobile application terminal to repair the network connection problem of the device calling without sound.

[0054] A second aspect of the present invention provides a screen status fault query and processing system, which is applied to any of the screen status fault query and processing methods, and specifically includes:

[0055] An identification module, the identification module is used to identify the screen status program sequence of the main playback device and the sub-playback device;

[0056] A fault query display module, which is used to display the screen status fault query result output by the mobile application terminal;

[0057] A graphic calculation module is used to calculate the area value of the negotiation area between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area.

[0058] The third aspect of the present invention provides a screen status fault query and processing terminal device, the screen status fault query and processing terminal device includes a memory and a processor, the memory stores a screen status fault query and processing method program, when the screen status fault query and processing method program is executed by the processor, any one of the steps of the screen status fault query and processing method is implemented.

[0059] The present invention solves the technical defects existing in the background technology, and the beneficial technical effects of the present invention are:

[0060] In the dynamic timing feature field, the screen state word vector changes after the first screen state program sequence and the second screen state program sequence are analyzed after dynamic timing update to obtain the first screen state; by obtaining the first future screen state timing model diagram and the second future screen state timing model diagram output in the target dynamic timing feature field when the main playback device is in the future playback period, the screen state output in the target dynamic timing feature field is analyzed to see whether it is synchronized, and the second screen state is obtained; if the screen state fault query result output by the mobile application terminal is displayed as the fault query result of the first screen state, the request code of the current partner API interface list is updated by calculating the difference between the current request response data and the historical request response data; if the screen state fault query result output by the mobile application terminal is displayed as the fault query result of the second screen state, an unqualified evaluation and assessment item is planned, and the target connection abnormal factors that cause the unqualified evaluation and assessment item are determined and repaired based on the unqualified evaluation and assessment item. The present invention can quickly and efficiently query and accurately solve the screen state faults between playback devices, thereby avoiding the phenomenon of templates without prices and calling numbers without sound between devices, and improving the synchronization and stability of screen playback of different devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0062] Figure 1 A first method flow chart of a screen status fault query and processing method is shown;

[0063] Figure 2 A second method flow chart of the screen status fault query and processing method is shown;

[0064] Figure 3 A system schematic diagram of a screen status fault query and processing system is shown;

[0065] Figure 4 A terminal device framework diagram of a screen status fault query and processing terminal device is shown. DETAILED DESCRIPTION

[0066] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0067] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0068] The first aspect of the present invention provides a screen status fault query and processing method, such as Figure 1 As shown, the following steps are included:

[0069] S102: Obtain a first screen state code sequence of a main playback device and a second screen state code sequence of a sub-playback device, analyze the screen state word vector changes of the first screen state code sequence and the second screen state code sequence after dynamic time sequence update in a dynamic time sequence feature field, and obtain a first screen state;

[0070] S104: acquiring a first future screen state timing model diagram and a second future screen state timing model diagram output in a target dynamic timing feature field when the main playback device is in a future playback period, so as to analyze whether the screen states output in the target dynamic timing feature field are synchronized, and obtain a second screen state;

[0071] S106: obtaining several types of joint screen non-playing faults that cause the first screen state and the second screen state to be generated between the main playback device and the sub-playback device, introducing the naive Bayes theorem to calculate and analyze the several types of joint screen non-playing faults, and deploying tests using natural language processing technology to obtain screen state fault query results;

[0072] S108: If the screen status fault query result output by the mobile application terminal is displayed as the fault query result of the first screen status, then after repairing the first screen status, the current request response data is obtained, and the difference between the current request response data and the historical request response data is calculated and analyzed to write a change modification calling code to update the request code of the current partner API interface list;

[0073] S110: If the screen status fault query result output by the mobile application terminal is displayed as the fault query result of the second screen status, unqualified assessment items are planned by constructing a network connection evaluation model and the normal status evaluation radar chart area corresponding to each assessment item, and the target connection abnormality factors that cause the unqualified assessment items are determined based on the unqualified assessment items and the best maintenance plan is used to repair them.

[0074] More specifically, the step S102 includes the following steps:

[0075] S202: Acquire a main playback device and one or more sub-playback devices interconnected with the main playback device, and identify a screen state program sequence of the main playback device, defined as a first screen state program sequence, and a screen state program sequence of the sub-playback device, defined as a second screen state program sequence, by using a mobile application terminal recognition technology;

[0076] S204: Acquire a text reading format of the mobile application terminal, and perform text conversion on the first screen state program sequence and the second screen state program sequence based on the text reading format to obtain first screen state text data and second screen state text data; wherein the mobile application terminal includes enterprise WeChat and a Wed end web page;

[0077] S206: introducing a stem feature extraction algorithm, performing word segmentation preprocessing on the first screen state text data and the second screen state text data, and performing stem feature extraction on the first screen state text data and the second screen state text data after the word segmentation preprocessing in the stem feature extraction algorithm to obtain a plurality of first screen state word vectors and a plurality of second screen state word vectors;

[0078] S208: constructing a dynamic temporal feature domain, presetting N mapping nodes in the dynamic temporal feature domain based on the plurality of first screen state word vectors and the plurality of second screen state word vectors, and planning a feature mapping sub-domain according to positions of the N mapping nodes in the dynamic temporal feature domain;

[0079] S210: Obtaining a dynamic time series interval corresponding to the feature mapping sub-field through a dynamic time series feature field, recording a screen state word vector after dynamic time series update on each mapping node based on the dynamic time series interval, and obtaining an updated first screen state word vector and an updated second screen state word vector;

[0080] S212: Introducing the cosine similarity method to calculate the degree of fit between the updated first screen state word vector and the updated second screen state word vector, if the degree of fit is less than a preset degree of fit, it indicates that there is a phenomenon that the template price is not displayed on a joint screen between the main playback device and the sub-playback device;

[0081] S214: Mark the connection status between the main playback device and the sub-playback device as a type of joint screen non-playback abnormal state and upload it to the mobile application terminal to generate a first screen state.

[0082] It should be noted that in certain specific application scenarios, it is usually necessary to use a movable or fixed sub-playback device to connect to the main playback device through the network to synchronize the playback content and interact in real time, such as using a mobile phone to connect to the restaurant's order preview terminal, so as to achieve the effect of joint screen ordering and price preview playback, which is convenient for the user's lifestyle. However, after the sub-playback device is connected to the main playback device, some prices in the template may not be displayed in the screen joint screen state failure. This may be caused by the two devices not being in the same local area network, using wireless connection, or the change of the partner API interface, which reduces the timing of the screen playback display. Therefore, in order to quickly and conveniently solve the above-mentioned fault problems, the method extracts the screen status word vector features of the main playback device and the sub-playback device through identification, and analyzes and calculates the screen status word vector features of the two after the dynamic timing interval update changes in the dynamic timing feature field to see whether the screen status word vector features outputted have a high degree of consistency. For example, the main device outputs the screen status of the main playback device after the dynamic timing interval changes, and the sub-device outputs the screen status of the sub-playback device after the same dynamic timing interval changes. If the screen status of the main playback device and the screen status of the sub-playback device have a consistency greater than the preset consistency in the dynamic timing feature field, it means that the screen status dynamic timing between the main playback device and the sub-playback device is synchronized, that is, the joint screen playback status is synchronized; otherwise, there is a joint screen failure problem, which makes the template price not displayed on the joint screen. On the one hand, this method can quickly analyze whether the screen-linked playback status between devices has a high degree of timing and synchronization, so as to accurately determine the template price-free failure problem between the screens of the devices, and provide a reliable analysis basis for optimizing the screen status; on the other hand, the extraction of screen status features can facilitate the analysis and processing of natural language technology, so that query processing can be carried out in the chat window of the mobile application terminal, further improving the convenience and speed of screen status fault query.

[0083] More specifically, the step S104 includes the following steps:

[0084] When there is no first screen state between the main playback device and the sub-playback device, the content playback plan and the current playback timestamp of the main playback device are obtained at this time;

[0085] Preset a future playback period, take the current playback timestamp as the starting time node, obtain multiple sets of first screen state text data of the main playback device in the future playback period based on the content playback plan and perform time sequence integration, and define it as a first future screen state text data set;

[0086] An ARIMA model algorithm is introduced to calculate the autoregressive order, the number of differences and the moving average order of the first future screen state text data set, and the autoregressive integrated moving average model is fitted through the autoregressive order, the number of differences and the moving average order to obtain the first future screen state time series dynamic model;

[0087] Setting a target dynamic timing node interval based on the current playback timestamp and the future playback timestamp, acquiring a target dynamic timing feature field corresponding to the target dynamic timing node region in the dynamic timing feature field, and dividing the target dynamic timing feature field into a plurality of uniform sub-fields;

[0088] Acquire the dynamic timing specifications of each sub-field in the dynamic timing feature field, embed the first future screen state timing dynamic model into the target dynamic timing feature field, uniformly time slice the first future screen state timing dynamic model according to the dynamic timing specifications of each sub-field, and obtain the first future screen state timing model diagram output by the main playback device in each sub-field;

[0089] Based on the above steps, the second future screen state timing model diagram output by the sub-playback device in each sub-field is calculated and obtained, and the first future screen state timing model diagram and the second future screen state timing model diagram corresponding to each sub-field are aligned. After the alignment is completed, the alignment coincidence degree between the first future screen state timing model diagram and the second future screen state timing model diagram in each sub-field is calculated;

[0090] Only sub-fields with a registration overlap less than a preset registration overlap are marked, which are defined as screen status abnormal sub-fields. The number of screen status abnormal sub-fields is counted. If the number of screen status abnormal sub-fields is greater than the preset number of sub-fields, it indicates that there is a phenomenon of no sound when the main playback device and the sub-playback device are connected to the screen;

[0091] The connection status between the main playback device and the sub-playback device is marked as a second-class joint screen non-playing state and uploaded to the mobile application terminal to generate a second screen state.

[0092] It should be noted that the dynamic timing specifications are descriptive parameters of state changes and event occurrences at different time points in the sub-field of the dynamic timing feature field. In addition to the above-mentioned template price-free phenomenon, in some application scenarios, there may also be a situation where there is no sound when calling the number on the joint screen. This is also caused by the failure of the sub-playback device to connect to the main playback device through the network to reach the content time period or the network connection between the devices, resulting in the failure of the joint screen playback or screenshot content between the devices to play, the content playback is not synchronized, and there is a delay. Therefore, the method can randomly slice and compare the dynamic screen state changes of the main playback device and the sub-playback device through the future playback period under the condition that there is no first screen state between the main playback device and the sub-playback device, that is, by constructing the future screen state timing model diagram output by the main playback device and the sub-playback device and performing timing slicing in the dynamic timing feature field, the sliced ​​timing model can intuitively freeze the screen state presented by each frame of the playback content on the same timeline between the devices, so as to ensure the high timing consistency of the real-time screen state comparison of each frame of the playback content between the main playback device and the sub-playback device, and accurately analyze the timing synchronization and delay of the screen state playback content of the two. Only the sub-fields with a registration overlap less than a preset registration overlap are marked, that is, the playback content with a time-series asynchronous screen state is extracted. If the number of these time-series asynchronous playback contents exceeds a certain preset number, it means that the multi-frame playback content in the same time sequence between the main playback device and the sub-playback device is inconsistent, that is, there are failure phenomena such as the playback content period not being reached and the playback delay, which may cause the joint screen call to be silent. This method can perform high-precision frame-by-frame analysis on the timing of playback content between devices, so as to quickly and accurately query whether there are playback content delays and asynchronous failures between the screen states of the devices, and improve the quality of joint screen playback of the device screens. Among them, the ARIMA model algorithm is called the autoregressive integrated moving average model in Chinese, which is a model for dynamic time series prediction. The ARIMA model algorithm combines the three core steps of autoregression, difference and moving average, and can effectively capture trends and seasonal changes in time series data.

[0093] More specifically, the step S106 includes the following steps:

[0094] Obtain the first historical screen state word vector corresponding to the first screen state between the main playback device and the sub-playback device, which is defined as the first historical word vector training set, and obtain the second historical screen state word vector corresponding to the second screen state between the main playback device and the sub-playback device, which is defined as the second historical word vector training set;

[0095] Acquire a network communication protocol between a main playback device and a sub-playback device, and retrieve and acquire, based on the network communication protocol, several types of joint screen non-playback failures that cause a first screen state and a second screen state to be generated between the main playback device and the sub-playback device in a big data network;

[0096] Count the occurrence frequencies of each type of joint screen non-playback failure in the first historical word vector training set and the second historical word vector training set, respectively, to obtain a plurality of first occurrence frequencies and second occurrence frequencies;

[0097] Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the first historical word vector training set according to the first occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the first screen state, which are defined as the first prior probability and the first conditional probability distribution;

[0098] Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the second historical word vector training set according to the second occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the second screen state, which are defined as the second prior probability and the second conditional probability distribution;

[0099] The naive Bayes theorem is introduced to calculate the first prior probability and the first conditional probability distribution, the second prior probability and the second conditional probability distribution, respectively, to obtain the first posterior probability and the second posterior probability, and only the joint screen non-playing fault type corresponding to the maximum first posterior probability and the maximum second posterior probability is extracted to generate the fault location query result;

[0100] A natural language processing model framework for the mobile application terminal is constructed through natural language processing technology, and the fault location query result is imported into the natural language processing model framework for deployment testing to obtain the recall rate. If the recall rate is lower than the preset recall rate, the fault location query result is re-predicted and adjusted until the recall rate reaches the preset recall rate to generate a screen status fault query result.

[0101] It should be noted that the types of failures in the joint screen not playing include the devices not being in the same local area network, the devices using wireless links, and the content playback time period not being reached. After querying the screen status failure between devices, different screen status failure problems are caused by different failure types and reasons. For example, the first screen status that appears between devices may be caused by the devices not being in the same local area network or changes in the partner's API interface. Therefore, different screen status failure problems will have multiple corresponding failure types. If the failure type cannot be accurately determined, it will be impossible to accurately handle and repair the existing screen status failure problem. This method extracts word vector features in the first screen state and in the second screen state as training sets to respectively train the fault type accurate positioning behavior under the two screen state fault problems; wherein, the occurrence frequency is the frequency of occurrence of each fault type in the training set, and this occurrence frequency calculation can be converted into the prior probability of which fault types cause the first screen state and the second screen state, which improves the reliability of fault type positioning training for different screen states compared with the traditional method that does not calculate the prior probability; and the conditional probability distribution is the probability of the characteristic frequency statistics and continuity distribution of each screen state word vector feature in the training set for each fault type. Through the conditional probability distribution, it can be ensured that the screen state fault problem accurately locates the fault type with a higher correlation, which provides an accurate and reliable fault tracing basis for the subsequent screen state fault problem processing and cause investigation. On the other hand, this method also introduces natural language processing technology to build a natural language processing model framework for mobile application terminals to deploy and test the trained fault location query results, so that fault problems can be queried and processed in a chat window and integrated into enterprise WeChat and web pages, thereby improving the convenience and efficiency of screen fault query processing. The recall rate is an indicator for evaluating the accuracy of the fault location query results evaluated by the natural language processing model test. A higher recall rate means that the fault location query results have higher accuracy and reliability, further ensuring the accuracy of fault type location.

[0102] More specifically, the step S108 includes the following steps:

[0103] If the screen state fault query result output by the mobile application terminal is displayed as a fault query result of the first screen state, then editing and generating fault avoidance prompt information according to the first screen state word vector and the second screen state word vector;

[0104] After the user repairs the first screen state between the main playback device and the sub-playback device according to the fault avoidance prompt information, reacquire the repaired first screen state word vector and the repaired second screen state word vector and calculate the degree of fit to obtain the repaired fit;

[0105] If the repaired fit is still less than the preset fit, construct an API request text parameter according to the repaired fit, obtain the current partner API interface list, send a request to the current partner API interface list based on the API request text parameter, and generate current request response data;

[0106] Preset a pre-repair historical update node, obtain historical request response data at the pre-repair historical update node, calculate and analyze the mean of the request path field between the current request response data and the historical request response data by variance analysis method, and obtain a mean square error value;

[0107] If the mean square error value is greater than the preset mean square error value, the deviation between the mean square error value and the preset mean square error value is calculated to obtain the mean square error deviation value, and at the same time, a list of partner API interfaces at the historical update node before the repair is obtained, which is defined as a historical partner API interface list;

[0108] Analyze the interface field differences between the current partner API interface list and the historical partner API interface list according to the mean square deviation value to obtain the partner API change content, and determine the calling code interval that needs to be modified in the current partner API interface list according to the partner API change content;

[0109] Edit the modified calling code based on the partner API change content, and apply the modified calling code to the calling code interval that needs to be modified to generate a new API calling document, update the request code of the current partner API interface list through the new API calling document, and generate updated request content;

[0110] The updated request content is uploaded to the mobile application terminal to automatically query and confirm and update the service request of the playback device in real time according to the change status of the partner API interface.

[0111] It should be noted that the fault query result of the first screen state displayed by the screen state fault query result can be used to check the fault problems existing in the first screen state one by one. For example, if the fault type that causes the template without price display between the main playback device and the sub-playback device is that the devices are not in the same local area network or the devices use a wireless link, then these fault types need to be automatically checked or manually set for processing. By processing these fault types, the phenomenon of no price display in the template between the devices may be repaired, that is, the degree of fit between the repaired first screen state word vector and the repaired second screen state word vector is greater than the preset fit. However, if the degree of fit between the screen state word vectors after the device is repaired is still less than the preset fit, it means that the phenomenon of no price display in the template for the joint screen playback between the devices still exists, indicating that these fault types are not the direct factors causing this phenomenon. This may be due to changes in the partner's API interface that cause the joint screen information service between the devices to be interrupted, data loss or errors, so that the templates of the two screens cannot display price information. Therefore, this method can send a verification request to the partner API interface list after the first screen state is repaired to verify whether the current screen access request is consistent with the partner API interface; wherein, the mean square error represents the difference between the current request response data and the historical request response data. If the mean square error value is greater than the preset mean square error value, it means that the difference between the current request response data and the historical request response data is large, which means that the partner API interface has been significantly updated or changed. In this way, it can provide an accurate analysis basis for determining and identifying the abnormal correlation between the screen state and the partner API change behavior. wherein, the request content includes request parameters, request methods and response processing; according to the detailed content of the API interface change, the part of the code that calls the old API is modified to adapt to the new API, and the request parameters, request methods and response processing are updated according to the new API call document, so that automatic query confirmation can be realized and the service request of the playback device can be updated in real time according to the change of the partner API interface, so that the screen request and the changed interface of the partner API maintain response stability and security, and solve and optimize the problem of the template not displaying price information and unstable service in the screen state of the playback device.

[0112] More specifically, the step S110 includes the following steps:

[0113] If the screen state fault query result output by the mobile application terminal is displayed as the fault query result of the second screen state, a current network connection node between the main playback device and the sub-playback device and a preset network communication channel on the current network connection node are obtained;

[0114] Based on the current network connection node, the preset network communication channel and the network communication protocol, a plurality of evaluation items for evaluating the network connection environment between the main playback device and the sub-playback device are obtained in the big data network, and a network connection evaluation model is created according to the plurality of evaluation items;

[0115] Build a radar chart architecture, obtain the lowest evaluation threshold of each evaluation item in a normal state based on the big data network, and plan the normal state evaluation radar chart area corresponding to each evaluation item in the radar chart architecture according to the lowest evaluation threshold;

[0116] A global evaluation is performed on the network connection environment between the current main playback device and the sub-playback device through a network connection evaluation model to obtain multiple evaluation indexes, and an actual evaluation radar chart is drawn in a radar chart framework based on the multiple evaluation indexes, and the negotiation area value between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area is calculated;

[0117] If the negotiation area value is smaller than the preset negotiation area value, the normal state assessment radar chart area is marked as an unqualified assessment item, and the assessment index corresponding to the unqualified assessment item is obtained;

[0118] Based on the big data network, all the connection abnormal factors that cause the network connection to produce unqualified evaluation items are obtained, and the fuzzy C-means algorithm is introduced. Based on the evaluation index, the clusters corresponding to each unqualified evaluation item are preset in the fuzzy C-means algorithm, and the membership matrix of each connection abnormal factor for each cluster is initialized;

[0119] Preset an iterative change threshold, calculate the center point of each cluster based on the membership matrix, obtain a number of cluster centers, continuously update the iterative membership matrix through the cluster centers, and obtain the membership matrix change after the update iteration. If the membership matrix change after the update iteration is less than the iterative change threshold, stop the iterative update, and finally output the target connection abnormality factor that causes the unqualified assessment item;

[0120] Obtain several processing solutions for target connection abnormality factors and the repair rate of each processing solution, extract the processing solution corresponding to the maximum repair rate and upload it to the mobile application terminal to repair the network connection problem of the device calling without sound.

[0121] It should be noted that, similarly, according to the fault query result of the second screen state displayed by the screen state fault query result, the fault problem existing in the second screen state is checked and repaired, and the network connection between the main playback device and the sub-playback device is evaluated and checked after the repair, but the traditional simple network connection evaluation method cannot accurately trace the specific factors that cause the abnormal network connection problem according to the evaluation results, so it is difficult to formulate a targeted maintenance plan for repair, which greatly reduces the timing and accuracy of the screen call between devices. This method first constructs a network connection evaluation model through several evaluation and assessment items of the network connection environment between the main playback device and the sub-playback device for global evaluation; wherein, the evaluation and assessment items include network connection quality, network connection rate, network connection stability and network connection environment; then constructs a radar chart framework, which is mainly used to visualize the qualified status of the network connection evaluation results and screen out the unqualified evaluation and assessment items of the network connection between devices in combination with the evaluation results, so as to improve the accuracy, traceability efficiency and reliability of tracing the abnormal network connection factors according to the evaluation results. Among them, if the radar charts of each evaluation and assessment item displayed in the actual evaluation radar chart negotiate with the corresponding normal state evaluation radar chart area, it means that the evaluation result of the evaluation and assessment item tends to be in a normal and stable state. Therefore, by further analyzing the area value of the negotiation area of ​​the two radar chart areas, it is possible to analyze which evaluation and assessment items are not within the normal and stable range, that is, unqualified evaluation and assessment items. Finally, based on the fuzzy C-means algorithm, it is calculated which connection abnormality factors these unqualified evaluation and assessment items belong to and adapted to the maintenance plan with a high repair rate to perform repairs; the fuzzy C-means algorithm can cluster the pointing effect of fuzzy and uncertain data, and can provide a smoother clustering transition for classes with overlapping features. At the same time, it has less impact on noise and outliers, and can more robustly identify data structures. Among them, the connection abnormality factors include modem failure, network adapter damage, network configuration errors, firewall blocking network access, network congestion, and wireless signal interference. This method can accurately trace the network connection abnormality factors that cause the screen joint screen calling failure, and adapt to efficient and targeted repair plans to perform processing, repair the network connection problem of the device calling without sound, and effectively improve the stability of screen joint screen playback.

[0122] In addition, the screen status fault query and processing method further includes the following steps:

[0123] Obtaining a partner API interface monitoring log of a target playback device, and extracting error records of the target playback device requesting a partner API interface within a preset time period through the partner API interface monitoring log;

[0124] Based on the error record, several request response error timestamps and the screen request API code sequence corresponding to each request response error timestamp are obtained, and the spatiotemporal kriging method is introduced to construct a dynamic spatiotemporal dimension map;

[0125] Calculating the variogram between each request response error timestamp and the corresponding screen request API code sequence by the spatiotemporal kriging method to obtain a plurality of variograms, and fitting the plurality of variograms in a dynamic spatiotemporal dimension graph to generate a dynamic spatiotemporal kriging equation;

[0126] Acquire the screen playback task of the target playback device, preset a future playback time node, extract the sub-playback task where the screen is at the future playback time node based on the screen playback task, and determine the screen request API sub-code where the screen is at the future playback time node according to the sub-playback task;

[0127] Solving the dynamic spatiotemporal Kriging equation based on the future playback time node, outputting a screen request API code sequence corresponding to a request response error at the future playback time node, which is defined as a screen request API code sequence of a future request exception;

[0128] Determine whether the screen request API subcode can be queried in the screen request API code sequence of the future request exception. If it can be queried, upload the screen request API subcode to the monitoring log in advance and update the screen request API subcode in advance to prevent errors when the screen requests the partner API interface at future time nodes.

[0129] It should be noted that when a user sends certain screen request APIs, if the partner API interface changes at this time, the screen request API sent by the user will cause error feedback, affecting the user's use and operation experience. This method can construct a dynamic space-time Kriging equation in the space-time Kriging method based on the request response error timestamp and the screen request API code sequence corresponding to each request response error timestamp. The space-time Kriging equation can accurately and dynamically predict the screen request API subcode that will cause the request response error at the future playback time node. Therefore, in order to avoid errors in the screen request API sent by the user, the screen request API code sent in the future should include this subcode; if the existence of this subcode has been queried, the corresponding subcode in the screen request API code sequence of the future request exception can be updated to adapt to the service request response requirements of the partner API change, to prevent the probability of frequent errors when the screen requests the partner API interface at future time nodes, and to improve the stability of the device screen connection.

[0130] In addition, the screen status fault query and processing method further includes the following steps:

[0131] Obtain the screen-linked playback status between the main playback device and the mobile playback device. If the screen-linked playback status is displayed as a non-playing state, obtain a network node diagram of the target network connected to the main playback device and the sub-playback device;

[0132] Constructing a network topology structure of a target network to which the main playback device and the sub-playback device are connected according to the network node diagram, and drawing a network wiring pattern diagram according to the network topology structure;

[0133] Based on the network wiring pattern diagram, all the transfer nodes in the network topology are marked, all the network communication transmission paths corresponding to the transfer nodes are planned with the transfer nodes as the radiation starting points, and the historical transmission error data corresponding to each network communication transmission path corresponding to each transfer node is obtained;

[0134] The Markov chain Monte Carlo algorithm is introduced to calculate the delay packet loss probability of the transit node for data transmission under the condition of the historical transmission error data corresponding to each network communication transmission path, and the delay packet loss probability of each transit node is obtained;

[0135] The transfer node with a calibrated delay packet loss probability greater than the preset delay packet loss probability is defined as a pending transfer node, the preset transmission parameters of the pending transfer node are obtained, and the data transmission weight value of each pending transfer node is calculated based on the preset transmission parameters;

[0136] If the data transmission weight value is less than the preset data transmission weight value, the pending transfer nodes corresponding to the data transmission weight values ​​less than the preset data transmission weight value are eliminated, and all network communication transmission paths corresponding to the remaining transfer nodes are obtained, which are defined as pending network communication transmission paths;

[0137] The Dijkstra algorithm is introduced. Based on the data transmission weight value of the remaining transfer nodes, the undetermined network communication transmission path corresponding to each transfer node is calculated and screened in the Dijkstra algorithm to generate the optimal network communication transmission path, and the optimal network communication transmission path is uploaded to the network management terminal of the mobile application.

[0138] It should be noted that the screen linkage between playback devices usually involves network connection, so it is crucial for the smoothness and efficiency of the network communication transmission structure. Complex or unreasonable network topology may cause data packets to pass through multiple transit nodes, increase the probability of delay and packet loss, and reduce the stability of the screen linkage of the playback device screen. Therefore, this method can screen out the transit nodes that have a high probability of delay and packet loss in the screen network signal data transmission and the network communication transmission path associated with this transit node, and select the best network communication transmission path from the network communication transmission path without delay and packet loss for signal data transmission, so as to improve the stability of signal transmission in the linkage state of the playback device screen, avoid the phenomenon of delayed packet loss causing the screen state to be unresponsive, and have high reliability. Among them, the Chinese name of the Dijkstra algorithm is the Dijkstra algorithm, which is used to screen the weighted single-source shortest path problem. The Dijkstra algorithm can ensure that the shortest path from the source node to all other nodes is found, so that the path screening result is accurate and reliable.

[0139] A second aspect of the present invention provides a screen status fault query and processing system, which is applied to any of the screen status fault query and processing methods described above, such as Figure 3 As shown, specifically including:

[0140] An identification module 301, which is used to identify the screen status program sequence of the main playback device and the sub-playback device;

[0141] A fault query display module 302, which is used to display the screen status fault query result output by the mobile application terminal;

[0142] The graphic calculation module 303 is used to calculate the area value of the negotiation region between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area.

[0143] The third aspect of the present invention provides a screen status fault query and processing terminal device, such as Figure 4 As shown, the screen status fault query and processing terminal device includes a memory 41 and a processor 42, and the memory 41 stores a screen status fault query and processing method program. When the screen status fault query and processing method program is executed by the processor 42, any one of the steps of the screen status fault query and processing method is implemented.

[0144] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A screen status fault query and processing method, characterized in that: The following steps are involved: S102: Obtain a first screen state program sequence of a main playback device and a second screen state program sequence of a sub-playback device, analyze the screen state word vector changes of the first screen state program sequence and the second screen state program sequence after dynamic time sequence update in the dynamic time sequence feature field, and obtain a first screen state; S104: acquiring a first future screen state timing model diagram and a second future screen state timing model diagram output in a target dynamic timing feature field when the main playback device is in a future playback period, so as to analyze whether the screen states output in the target dynamic timing feature field are synchronized, and obtain a second screen state; S106: obtaining several types of joint screen non-playing faults that cause the first screen state and the second screen state to be generated between the main playback device and the sub-playback device, introducing the naive Bayes theorem to calculate and analyze the several types of joint screen non-playing faults, and deploying tests using natural language processing technology to obtain screen state fault query results; S108: If the screen status fault query result output by the mobile application terminal is displayed as the fault query result of the first screen status, then after repairing the first screen status, the current request response data is obtained, and the difference analysis between the current request response data and the historical request response data is calculated to edit and change the modification calling code to update the first screen status request code of the partner API interface list; S110: If the screen status fault query result output by the mobile application terminal is displayed as a fault query result of the second screen status, unqualified assessment items are planned by constructing a network connection assessment model and a normal status assessment radar chart area corresponding to each assessment item, and based on the unqualified assessment item, the target connection abnormality factor causing the second screen status is determined and repaired using the best maintenance solution; The step S102 specifically includes the following steps: Obtain a main playback device and one or more sub-playback devices interconnected with the main playback device, and identify a screen state program sequence of the main playback device, defined as a first screen state program sequence, and a screen state program sequence of the sub-playback device, defined as a second screen state program sequence, through a mobile application terminal recognition technology; Acquire a text reading format of the mobile application terminal, and perform text conversion on the first screen state program sequence and the second screen state program sequence based on the text reading format to obtain first screen state text data and second screen state text data; wherein the mobile application terminal includes enterprise WeChat and a web page; Introducing a stem feature extraction algorithm, performing word segmentation preprocessing on the first screen state text data and the second screen state text data, and performing stem feature extraction on the first screen state text data and the second screen state text data after the word segmentation preprocessing in the stem feature extraction algorithm to obtain a plurality of first screen state word vectors and a plurality of second screen state word vectors; Constructing a dynamic time series feature domain, presetting N mapping nodes in the dynamic time series feature domain based on the plurality of first screen state word vectors and the plurality of second screen state word vectors, and planning a feature mapping sub-domain according to positions of the N mapping nodes in the dynamic time series feature domain; Obtaining a dynamic time series interval corresponding to the feature mapping sub-field through a dynamic time series feature field, recording a screen state word vector after dynamic time series update on each mapping node based on the dynamic time series interval, and obtaining an updated first screen state word vector and an updated second screen state word vector; The cosine similarity method is introduced to calculate the degree of fit between the updated first screen state word vector and the updated second screen state word vector. If the degree of fit is less than the preset degree of fit, it indicates that there is a phenomenon that the template price is not displayed on a joint screen between the main playback device and the sub-playback device. The screen linkage state between the main playback device and the sub-playback device is marked as a type of linked screen non-playback abnormal state and uploaded to the mobile application terminal to generate a first screen state.

2. The screen status fault query and processing method according to claim 1, characterized in that: The step S104 specifically includes the following steps: When there is no first screen status fault query result between the main playback device and the sub-playback device, then the content playback plan and the current playback timestamp of the main playback device are obtained; Preset a future playback period, take the current playback timestamp as the starting time node, obtain multiple sets of first screen state text data of the main playback device in the future playback period based on the content playback plan and perform time sequence integration, and define it as a first future screen state text data set; An ARIMA model algorithm is introduced to calculate the autoregressive order, the number of differences and the moving average order of the first future screen state text data set, and the autoregressive integrated moving average model is fitted through the autoregressive order, the number of differences and the moving average order to obtain the first future screen state time series dynamic model; Setting a target dynamic timing node interval based on the current playback timestamp and the future playback timestamp, acquiring a target dynamic timing feature field corresponding to the target dynamic timing node region in the dynamic timing feature field, and dividing the target dynamic timing feature field into a plurality of uniform sub-fields; Acquire the dynamic timing specifications of each sub-field in the dynamic timing feature field, embed the first future screen state timing dynamic model into the target dynamic timing feature field, uniformly time slice the first future screen state timing dynamic model according to the dynamic timing specifications of each sub-field, and obtain the first future screen state timing model diagram output by the main playback device in each sub-field; Based on the above steps, the second future screen state timing model diagram output by the sub-playback device in each sub-field is calculated and obtained, and the first future screen state timing model diagram and the second future screen state timing model diagram corresponding to each sub-field are aligned. After the alignment is completed, the alignment coincidence degree between the first future screen state timing model diagram and the second future screen state timing model diagram in each sub-field is calculated; Only sub-fields with a registration overlap less than a preset registration overlap are marked, which are defined as screen status abnormal sub-fields. The number of screen status abnormal sub-fields is counted. If the number of screen status abnormal sub-fields is greater than the preset number of sub-fields, it indicates that there is a phenomenon of no sound when the main playback device and the sub-playback device are connected to the screen; The screen linkage state between the main playback device and the sub-playback device is marked as a second-class linked screen non-playing state and uploaded to the mobile application terminal to generate a second screen state.

3. The screen status fault query and processing method according to claim 1, characterized in that: The step S106 specifically includes the following steps: Obtain the first historical screen state word vector corresponding to the first screen state between the main playback device and the sub-playback device, which is defined as the first historical word vector training set, and obtain the second historical screen state word vector corresponding to the second screen state between the main playback device and the sub-playback device, which is defined as the second historical word vector training set; Acquire a network communication protocol between a main playback device and a sub-playback device, and retrieve and acquire, based on the network communication protocol, several types of joint screen non-playback failures that cause a first screen state and a second screen state to be generated between the main playback device and the sub-playback device in a big data network; Count the occurrence frequencies of each type of joint screen non-playback failure in the first historical word vector training set and the second historical word vector training set, respectively, to obtain a plurality of first occurrence frequencies and second occurrence frequencies; Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the first historical word vector training set according to the first occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the first screen state, which are defined as the first prior probability and the first conditional probability distribution; Calculate the prior probability and conditional probability layout of each type of unplayed fault on the joint screen and each word vector feature in the second historical word vector training set according to the second occurrence frequency, and obtain the prior probability and conditional probability distribution of each type of unplayed fault on the joint screen relative to the second screen state, which are defined as the second prior probability and the second conditional probability distribution; The naive Bayes theorem is introduced to calculate the first prior probability and the first conditional probability distribution, the second prior probability and the second conditional probability distribution, respectively, to obtain the first posterior probability and the second posterior probability, and only the joint screen non-playing fault type corresponding to the maximum first posterior probability and the maximum second posterior probability is extracted to generate the fault location query result; A natural language processing model framework for the mobile application terminal is constructed through natural language processing technology, and the fault location query result is imported into the natural language processing model framework for deployment testing to obtain the recall rate. If the recall rate is lower than the preset recall rate, the fault location query result is re-predicted and adjusted until the recall rate reaches the preset recall rate to generate a screen status fault query result.

4. The screen status fault query and processing method according to claim 1, characterized in that: The step S108 specifically includes the following steps: If the screen state fault query result output by the mobile application terminal is displayed as a fault query result of the first screen state, then editing and generating fault avoidance prompt information according to the first screen state word vector and the second screen state word vector; After the user repairs the first screen state between the main playback device and the sub-playback device according to the fault avoidance prompt information, reacquire the repaired first screen state word vector and the repaired second screen state word vector and calculate the degree of fit to obtain the repaired fit; If the repaired fit is still less than the preset fit, construct an API request text parameter according to the repaired fit, obtain the current partner API interface list, send a request to the current partner API interface list based on the API request text parameter, and generate current request response data; Preset a pre-repair historical update node, obtain historical request response data at the pre-repair historical update node, calculate and analyze the mean of the request path field between the current request response data and the historical request response data by variance analysis method, and obtain a mean square error value; If the mean square error value is greater than the preset mean square error value, the deviation between the mean square error value and the preset mean square error value is calculated to obtain the mean square error deviation value, and at the same time, a list of partner API interfaces at the historical update node before the repair is obtained, which is defined as a historical partner API interface list; Analyze the interface field differences between the current partner API interface list and the historical partner API interface list according to the mean square deviation value to obtain the partner API change content, and determine the calling code interval that needs to be modified in the current partner API interface list according to the partner API change content; Edit the modified calling code based on the partner API change content, and apply the modified calling code to the calling code interval that needs to be modified to generate a new API calling document, update the request code of the current partner API interface list through the new API calling document, and generate updated request content; The updated request content is uploaded to the mobile application terminal to automatically query and confirm and update the service request of the playback device in real time according to the change status of the partner API interface.

5. The screen status fault query and processing method according to claim 1, characterized in that: The step S110 specifically includes the following steps: If the screen state fault query result output by the mobile application terminal is displayed as the fault query result of the second screen state, a current network connection node between the main playback device and the sub-playback device and a preset network communication channel on the current network connection node are obtained; Based on the current network connection node, the preset network communication channel and the network communication protocol, a plurality of evaluation items for evaluating the network connection environment between the main playback device and the sub-playback device are obtained in the big data network, and a network connection evaluation model is created according to the plurality of evaluation items; Build a radar chart architecture, obtain the lowest evaluation threshold of each evaluation item in a normal state based on the big data network, and plan the normal state evaluation radar chart area corresponding to each evaluation item in the radar chart architecture according to the lowest evaluation threshold; A global evaluation is performed on the network connection environment between the current main playback device and the sub-playback device through a network connection evaluation model to obtain multiple evaluation indexes, and an actual evaluation radar chart is drawn in a radar chart framework based on the multiple evaluation indexes, and the negotiation area value between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area is calculated; If the negotiation area value is smaller than the preset negotiation area value, the normal state assessment radar chart area is marked as an unqualified assessment item, and the assessment index corresponding to the unqualified assessment item is obtained; Based on the big data network, all the connection abnormal factors that cause the network connection to produce unqualified evaluation items are obtained, and the fuzzy C-means algorithm is introduced. Based on the evaluation index, the clusters corresponding to each unqualified evaluation item are preset in the fuzzy C-means algorithm, and the membership matrix of each connection abnormal factor for each cluster is initialized; Preset an iterative change threshold, calculate the center point of each cluster based on the membership matrix, obtain a number of cluster centers, continuously update the iterative membership matrix through the cluster centers, and obtain the membership matrix change after the update iteration. If the membership matrix change after the update iteration is less than the iterative change threshold, stop the iterative update, and finally output the target connection abnormality factor that causes the unqualified assessment item; Obtain several processing solutions for target connection abnormality factors and the repair rate of each processing solution, extract the processing solution corresponding to the maximum repair rate and upload it to the mobile application terminal to repair the network connection problem of the device calling without sound.

6. A screen status fault query and processing system, applied to the screen status fault query and processing method according to any one of claims 1 to 5, characterized in that: Specifically include: An identification module, the identification module is used to identify the screen status program sequence of the main playback device and the sub-playback device; A fault query display module, which is used to display the screen status fault query result output by the mobile application terminal; A graphic calculation module is used to calculate the area value of the negotiation area between each evaluation item radar chart displayed in the actual evaluation radar chart and the corresponding normal state evaluation radar chart area.

7. A terminal device for querying and processing screen status faults, characterized in that: The terminal device for screen status fault query and processing includes a memory and a processor, wherein a screen status fault query and processing method program is stored in the memory. When the screen status fault query and processing method program is executed by the processor, the screen status fault query and processing method steps as described in any one of claims 1 to 5 are implemented.

Images