Method and system for predicting inspection data of operation and maintenance equipment, equipment and medium

Abstract

The invention belongs to the technical field of railway operation and maintenance, and discloses an operation and maintenance equipment inspection data prediction method, system, equipment and medium, and the method comprises the steps: obtaining first historical inspection data of a detection item corresponding to operation and maintenance equipment; preprocessing the first historical inspection data of the detection item to obtain second historical inspection data related to inspection times; based on the second historical inspection data and an inspection data state transition model, determining state transition reference inspection data of the detection item; based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix, constructing a state transition probability prediction model; and predicting the current inspection data in the detection item by using the state transition probability prediction model to obtain the probability of transition of the current inspection data to the next state. According to the scheme of the invention, the inspection plan of a fixed shift can be converted into inspection according to the state, and the density of inspection work can be effectively reduced.

Description

Technical Field

[0001] This application relates to the technical field of railway operation and maintenance, and particularly to a prediction method, system, device and medium for inspection data of operation and maintenance equipment. Background Art

[0002] The maintenance of railway equipment is an important task to ensure the safety of railway operation. To ensure the quality of equipment maintenance, a large number of maintenance tasks are arranged. Such maintenance tasks include related equipment in all aspects of the railway system. For example, cleaning the surface of the equipment, swapping device cards, and conducting station shutdown tests. The current railway maintenance work already covers all aspects of the railway system.

[0003] However, this way of ensuring the maintenance quality by presetting the maintenance plan will actually result in over-maintenance. The maintenance of equipment is actually not lossless. For example, when wiping the dust on the surface of the equipment, some switches may be accidentally touched. Once such unexpected situations outside the operation and maintenance plan occur, it may lead to the inability to handle the operation and maintenance work within the planned time. It will affect the overall transportation efficiency of the railway at a minimum and endanger the train operation safety at a maximum. Summary of the Invention

[0004] To solve the above problems, a prediction method, system, device and medium for inspection data of operation and maintenance equipment provided by this application can transform the inspection plan from a fixed-shift inspection plan into an inspection based on the status, which can effectively reduce the density of inspection work.

[0005] To achieve the above object, this application adopts the following technical solutions:

[0006] In the first aspect, the present invention provides a prediction method for inspection data of operation and maintenance equipment, including:

[0007] Obtain the first historical inspection data of the detection items corresponding to the operation and maintenance equipment;

[0008] Preprocess the first historical inspection data of the detection items to obtain the second historical inspection data related to the number of inspections;

[0009] Based on the second historical inspection data and the inspection data state transition model, determine the state transition reference inspection data of the detection items;

[0010] Based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix, construct a state transition probability prediction model;

[0011] Use the state transition probability prediction model to predict the current inspection data in the detection items to obtain the probability of the current inspection data transitioning to the next state.

[0012] In the technical solution provided by this application, after the operation and maintenance equipment completes one inspection, it can obtain the probability of the detection data transferring to the next state. Therefore, when constructing the inspection plan, there is no need to set the inspection plan according to a fixed plan. The inspection plan can be transformed from a fixed shift inspection plan to an inspection based on the state. In this way, the density of the inspection work can be effectively reduced, and at the same time, it is ensured that the inspection work can timely detect potential safety hazards.

[0013] Further, preprocess the first historical inspection data of the inspection item, including:

[0014] Normalize the discrete data obtained by each active detection in the first historical inspection data; and / or

[0015] Divide the continuous data in the first historical inspection data into intervals, and convert the data within the intervals using a conversion coefficient; and / or

[0016] For the state data in the first historical inspection data, record the duration of each state corresponding to each state switch.

[0017] In the solution provided by this application, the historical inspection data includes discrete data, continuous data or state data. For these three different types of data, three different methods are used for corresponding preprocessing. In this way, the characteristics of the corresponding inspection data can be reflected, and then the development direction of the inspection data under the corresponding inspection item can be accurately distinguished, and it provides convenience for establishing the corresponding state transition matrix in the future.

[0018] Further, based on the second historical inspection data and the inspection data state transition model, determine the state transition reference inspection data of the inspection item, including:

[0019] Accumulate the first data point in the second historical inspection data to the current data point to obtain the third historical inspection data;

[0020] Based on the second historical inspection data and the third historical inspection data, construct an inspection data state transition model;

[0021] Based on the unbiased parameters of the second historical inspection data and the inspection data state transition model, determine the state transition reference inspection data of the inspection item.

[0022] In the solution provided by this application, when generating the state transition reference inspection data, on the basis of limited data, new sequences can be generated by accumulation, and the parameters of the inspection data state transition model can be solved by using the adjacent mean sequence and the least squares method, so as to obtain a state transition probability with a higher correlation with each data point in the historical inspection data, and then increase the information contained in each data point in the historical inspection data, so as to have better accuracy when predicting the state transition probability further in the future.

[0023] Further, based on the second historical inspection data, the state transition reference inspection data, and the state transition probability matrix, a state transition probability prediction model is constructed, including:

[0024] Establish a deviation value sequence between the second historical inspection data and the state transition reference inspection data;

[0025] Based on the deviation value sequence and different thresholds or ranges, divide the inspection items into several inspection states;

[0026] Based on several inspection states, establish a state transition probability matrix;

[0027] Based on the median of the state space of each inspection state among several inspection states, establish a state median matrix;

[0028] Based on the state transition reference inspection data, the state transition probability matrix, and the state median matrix, construct a state transition probability prediction model.

[0029] In this solution, the inspection items are divided into multiple time states from the healthy state to the maintenance state, so the entire process of the inspection object from normal operation to wear and tear and then to damage can be accurately described. Furthermore, corresponding state thresholds can be set in practice to measure the maintenance plan. By combining the historical inspection data with the state transition probability matrix and the state median matrix, a state transition probability prediction model can be constructed to accurately predict the probability of the inspection transitioning to the next state. Therefore, when constructing the inspection plan, instead of setting the inspection plan according to a fixed plan, the inspection plan can be transformed from a fixed shift inspection plan to an inspection based on the state.

[0030] In the second aspect, the present invention also provides a prediction system for the inspection data of operation and maintenance equipment, including:

[0031] An acquisition module, configured to acquire the first historical inspection data of the inspection items corresponding to the operation and maintenance equipment;

[0032] A processing module, configured to preprocess the first historical inspection data of the inspection items to obtain the second historical inspection data related to the number of inspections;

[0033] A determination module, configured to determine the state transition reference inspection data of the inspection items based on the second historical inspection data and the inspection data state transition model;

[0034] A construction module, configured to construct a state transition probability prediction model based on the second historical inspection data, the state transition reference inspection data, and the state transition probability matrix;

[0035] A prediction module, configured to predict the current inspection data in the inspection item by using a state transition probability prediction model, and obtain the probability of the current inspection data transitioning to the next state.

[0036] Further, a processing module is further configured to:

[0037] Normalize the discrete data obtained by each active detection in the first historical inspection data; and / or

[0038] Divide the continuous data in the first historical inspection data into intervals, and convert the data within the intervals by using a conversion coefficient; and / or

[0039] Record the duration corresponding to each state at each state transition for the state data in the first historical inspection data.

[0040] Further, a determination module is further configured to:

[0041] Accumulate the first data point in the second historical inspection data to the current data point to obtain the third historical inspection data;

[0042] Construct an inspection data state transition model based on the second historical inspection data and the third historical inspection data;

[0043] Determine the state transition reference inspection data of the inspection item based on the second historical inspection data and the unbiased parameters of the inspection data state transition model.

[0044] Further, a construction module is further configured to:

[0045] Establish a deviation value sequence between the second historical inspection data and the state transition reference inspection data;

[0046] Divide the inspection item into several inspection states based on the deviation value sequence and different thresholds or ranges;

[0047] Establish a state transition probability matrix based on the several inspection states;

[0048] Establish a state median matrix based on the median of the state space of each inspection state in the several inspection states;

[0049] Construct a state transition probability prediction model based on the state transition reference inspection data, the state transition probability matrix, and the state median matrix.

[0050] In a third aspect, the present invention further provides an electronic device, including: a processor and a memory;

[0051] The processor is coupled to the memory;

[0052] Among them, the processor is used to read and execute the programs or instructions stored in the memory, so that the device executes the method as described in the first aspect.

[0053] In a fourth aspect, the present invention further provides a computer-readable storage medium storing a computer program, and when the program is executed by a processor, the method as described in the first aspect is implemented.

[0054] The technical solution provided by this application has at least the following technical effects or advantages:

[0055] In the technical solution provided by this application, the historical inspection data of the corresponding inspection items of the operation and maintenance device is preprocessed to obtain historical inspection data related to the number of inspections, combined with the state transition reference inspection data of the whitening differential equation inspection items, and then combined with the state transition probability matrix and the state median matrix to construct a state transition probability prediction model. The state transition probability prediction model is used to predict the current inspection data in the inspection items to obtain the probability of the current inspection data transitioning to the next state. In the technical solution of this application, after the operation and maintenance device completes an inspection, the probability of the inspection data transitioning to the next state can be obtained. Therefore, when constructing the inspection plan, there is no need to set the inspection plan according to a fixed plan. The inspection plan can be transformed from a fixed-shift inspection plan to an inspection based on the state. Furthermore, the density of the inspection work can be effectively reduced, while ensuring that the inspection work can timely detect potential safety hazards.

[0056] Other features and advantages of this application will be described in the subsequent description, and some of them will become obvious from the description or be understood by implementing this application. The objectives and other advantages of this application can be achieved and obtained through the structures pointed out in the description, claims, and drawings. Description of the Drawings

[0057] In order to more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of this application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0058] Figure 1 It is a schematic flowchart of a method for predicting inspection data of an operation and maintenance device in an embodiment of this application;

[0059] Figure 2 It is a schematic structural diagram of a system for predicting inspection data of an operation and maintenance device in an embodiment of this application;

[0060] Figure 3 It is a schematic structural diagram of an electronic device provided in an embodiment of this application. Detailed implementation manners

[0061] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without making creative efforts shall fall within the protection scope of the present application.

[0062] Figure 1 As shown in the flowchart of a method for predicting inspection data of an operation and maintenance device in an embodiment of the present application, the method includes:

[0063] S101. Obtain the first historical inspection data of the inspection items corresponding to the operation and maintenance device;

[0064] In the present application, the inspection object refers to a device or apparatus that can detect data through manual or existing detection technologies, such as devices related to the use of trains in the railway system, including: tracks, vehicle brake pads, temperature sensors in specific areas, smoke alarms, wheels, etc. The data obtained by detecting the inspection object, that is, the operation and maintenance data of the detection object, such as the data on the wear condition of the track surface, the data on the thickness of the vehicle brake pads, the temperature in a specific area, the usage condition of the smoke alarm, the wear condition of the wheels, etc. The operation and maintenance data cannot be data related to rules or regulations, such as the train running route, the length of the train carriage, etc.

[0065] In the present application, the operation and maintenance device refers to an inspection device, that is, an inspection device used in the operation and maintenance work, which is used to input the operation and maintenance data of the detection object. The operation and maintenance data input into the inspection device can also be called inspection data. The inspection data contains the operation and maintenance data obtained by detecting the detection object and has a time attribute. Therefore, the inspection data can be regarded as data conforming to the time series. The inspection item refers to the complete process of inspecting the detection object from the healthy state to the maintenance state. The data set of the operation and maintenance data of the detection object input into the inspection device from the healthy state to the maintenance state is the first historical inspection data of the inspection item.

[0066] S102. Preprocess the first historical inspection data of the inspection item to obtain the second historical inspection data related to the number of inspections;

[0067] In the specific detection items, the first historical inspection data includes continuous data, discrete data, or status data. Continuous data refers to the operation and maintenance data obtained by real-time monitoring of the detection object, such as the driving mileage of a train. Discrete data refers to the operation and maintenance data that can only be obtained during each active detection, such as the wear condition of the track. Status data refers to the operation and maintenance data with specific status values, such as the on state and off state of the emergency lights in the railway system, corresponding to the two states of on and off.

[0068] Normalize the discrete data obtained from each active detection in the first historical inspection data, that is, divide the discrete data obtained from each active detection into the same interval range for easy comparison and calculation. The value corresponding to each interval after normalization can be regarded as the data point corresponding to one inspection.

[0069] Divide the continuous data obtained from real-time monitoring in the first historical inspection data into intervals, and convert the data within the interval into corresponding values using a conversion coefficient. For example, select several intervals at intervals and take the median within each interval. The value obtained by converting the data within the interval using the conversion coefficient can be regarded as the data point corresponding to one inspection.

[0070] For the status data in the first historical inspection data, record the duration of each state corresponding to each state switch. The value of each state and the duration of each state can be regarded as the data point corresponding to one inspection. Exemplarily, the data points corresponding to the inspection of the standby power supply are working for 1 hour, sleeping for 20 hours, working for 1 hour, etc., which are represented by two numbers. 0 represents the working state, and 1 represents the sleeping state. For example, (0,1) means working for 1 hour, and (1,20) means sleeping for 20 hours.

[0071] After preprocessing the continuous data, discrete data, or status data in the first historical inspection data of the detection item, the inspection data sequence composed of the data points corresponding to each inspection is obtained, which is the second historical inspection data related to the number of inspections.

[0072] S103. Determine the status transfer reference inspection data of the detection item based on the second historical inspection data and the inspection data state transfer model;

[0073] The second historical inspection data can be expressed as X (0) , X (0) = {x (0) (1), x (0) (2), x (0) (3)……x (0) (n)}; where n represents the number of data points in the historical inspection data, and x (0) (n) represents the nth data point in the second historical inspection data X (0) ;

[0074] For each data point in the second historical inspection data, accumulate the first data point in the second historical inspection data to the current data point to obtain the cumulative sequence X (1) , and this cumulative data can be called the third historical inspection data, which can be expressed as X (1) = {x (1) (1), x (1) (2), x (1) (3)…x (1) (i)…x (1) (n)};

[0075] The calculation formula for the data points in the cumulative sequence can be expressed as:

[0076]

[0077] where i is the index of the data point, and x (0) (i) represents the i-th data point in the second historical inspection data X (0) , and the number of data points in the cumulative sequence X (1) is equal to the number of data points in the second historical inspection data X (0) .

[0078] Based on the second historical inspection data and the third historical inspection data, construct an inspection data state transition model;

[0079] Based on the third historical inspection data, establish a whitenization differential equation, expressed as:

[0080]

[0081] where q and Q are the parameters of the whitenization differential equation, q is the development coefficient, Q is the grey action quantity, and d is the differential symbol.

[0082] Based on the mean value of adjacent data points in the third historical inspection data, generate an adjacent sequence corresponding to the third historical inspection data, and this adjacent sequence can be called the fourth historical inspection data, expressed as X (2) , X (2) = {x (2) (1), x (2) (2), x (2) (3)…x (2) (u)…x (2) (n)};

[0083] where u represents the index of the data point in the historical inspection data.

[0084]

[0085] The calculation formula for the data points in the adjacent sequence can be expressed as:

[0086]

[0087] Based on the fourth historical inspection data and the second historical inspection data, determine the first intermediate parameter A and the second intermediate parameter B;

[0088] Construct the intermediate parameter matrix A and the vector B:

[0089] The matrix A is an (n - 1)×2 matrix, where the first column is -x (2) (u) (from u = 2 to u = n), and the second column is all 1s;

[0090] The vector B is an (n - 1)-dimensional column vector, and its elements are x (0) (u) (from u = 2 to u = n);

[0091]

[0092] Based on the first intermediate parameter A and the second intermediate parameter B, simplify the whitening differential equation. The simplified whitening differential equation can be expressed as:

[0093] d = (A T A) -1 A T B;

[0094] where d is the unbiased parameter of q, T is the matrix transformation symbol, A T is the transpose matrix of A, (A T A) -1 is the inverse matrix of (A T A).

[0095] Solve the simplified whitening differential equation to obtain the development coefficient and the grey action amount of the whitening differential equation; solve the simplified whitening differential equation by the least squares method to obtain the unbiased parameter v.

[0096] The relational expression between the unbiased parameter v and the development coefficient q is:

[0097]

[0098] Solving the equation can obtain the development coefficient q of the whitening differential equation, and the grey action amount Q can be obtained through the development coefficient q;

[0099] The relational expression between the unbiased parameter V and the grey action amount Q is:

[0100]

[0101] Solving the equation can obtain the unbiased parameter V.

[0102] The differential equation for determining the development coefficient and the grey action quantity is determined as the state transition model of the inspection data;

[0103] Based on the unbiased parameters of the second historical inspection data and the state transition model of the inspection data, the state transition reference inspection data of the inspection item is determined, which can be expressed as:

[0104]

[0105] where e is the natural constant and k is the index of the data point of the state transition reference inspection data X (3) of the data.

[0106] The state transition reference inspection data is actually the numerical change situation of the historical inspection data characterized after the historical inspection data in the inspection item are combined together.

[0107] S104. Based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix, construct a state transition probability prediction model;

[0108] Establish a deviation value sequence H between the second historical inspection data and the state transition reference inspection data, which is expressed as: H = {h(1), h(2)…h(g)…h(n - 1)}, where g is the index of the data point in the deviation value sequence H and h ∈ [1, n - 1];

[0109] The calculation formula for the data point in the deviation value sequence can be expressed as:

[0110] h(g) = x (0) (g) - x (3) (g);

[0111] Based on the deviation value sequence and different thresholds or ranges, divide the inspection item into several inspection states;

[0112] In this application, the inspection state refers to dividing the whole process of the inspection item from the healthy state to the maintenance state into several time tenses by setting different thresholds or ranges. Each time tense corresponds to an inspection state, and the state space of each inspection state corresponds to one or more inspection data points. The inspection state can be determined according to the specific inspection item, that is, the inspection state is determined according to the specific inspection object.

[0113] Taking the deviation value sequence as the base value, that is, taking the data points in the deviation value sequence H as the basis for dividing the corresponding inspection state of the inspection item. By setting different thresholds or ranges, the inspection item is divided into several inspection states.

[0114] Assume that the second historical inspection data X (0) is {80, 82, 85, 88, 90}, and the state transition reference inspection data X (3)If it is {81, 83, 84, 87, 89}, then the deviation value sequence H is {80 - 81, 82 - 83, 85 - 84, 88 - 87, 90 - 89} = {-1, -1, 1, 1, 1}.

[0115] Divide the detection status based on the deviation value sequence as the base value:

[0116] Set that the deviation value between [-1, 1] is the "normal state", the deviation value less than -1 is the "low state", and the deviation value greater than 1 is the "high state".

[0117] According to the divided detection status, the elements -1 and -1 in the deviation value sequence H correspond to the "normal state", and 1, 1, and 1 correspond to the "high state".

[0118] Based on the divided several detection statuses, establish a state transition probability matrix P, which can be expressed as:

[0119]

[0120] Among them, m represents the number of detection statuses, and p 1,1 represents the probability of transferring from the first detection status to the first detection status, and p 1,m represents the probability of transferring from the 1st detection status to the mth detection status, and p m,1 represents the probability of transferring from the mth detection status to the 1st detection status, and p m,m represents the probability of the mth detection status transferring to the mth detection status; the value of p is the proportion of the number of transfers from detection status 1 to detection status 1 in all transfers of detection status 1 in the historical patrol inspection data, which is directly obtained through frequency statistics.

[0121] Exemplarily, the detection status sequence of the historical patrol inspection data is [1, 1, 2, 1, 1, 3, 1], then the state transition statistics are as follows:

[0122] N 1,1 = 3 (1→1, 1→1, 1→1)

[0123] N 1,2 = 1 (1→2)

[0124] N 1,3 = 1 (1→3)

[0125] The total number of transfers starting from state 1 is 3 + 1 + 1 = 5. Therefore:

[0126]

[0127] Based on the median of the state space of each detection status among several detection statuses, establish a state median matrix W T , which can be expressed as: W​​T = [w1, w2, … w m ; where w1 represents the median of the state space of the first detection state, w2 represents the median of the state space of the second detection state, and w m represents the median of the state space of the m-th detection state.

[0128] Based on the state transition benchmark inspection data, the state transition probability matrix, and the state median matrix, a state transition probability prediction model is constructed, which can be expressed as:

[0129] δ`(y + r) = x (3) (y + r) + p r + P × W;

[0130] where y represents the index of the first data point where a detection state transition occurs in the second historical inspection data, which can also be called the state transition starting point; x (3) (y + r) represents the benchmark value of the (y + r)-th prediction point; δ`(y + r) represents the predicted value of the (y + r)-th prediction point, that is, the probability of the prediction point transitioning to the next state; p r represents the state probability vector of the (y + r)-th prediction point, r represents the number of data points from the starting point at the r-th data point, W represents the state median matrix, and P represents the state transition probability matrix.

[0131] Since the second historical inspection data for constructing the state transition probability prediction model is a sequence related to the inspection times, the index of the data point in the sequence represents the inspection times corresponding to the data point, and the state transition starting point y corresponds to the y-th inspection. r represents the inspection times after the y-th inspection, and the predicted value δ`(y + r) represents the probability of the data point of the (r + y)-th inspection transitioning to the next state. After the state transition probability prediction model is determined, the starting point y value is a determined value. When predicting the current inspection data, the difference between the inspection times corresponding to the current inspection data and the state transition starting point y is the inspection times r after the y-th inspection. Inputting the obtained inspection times r into the state transition probability prediction model can obtain the probability δ`(y + r) of the current inspection data transitioning to the next state.

[0132] In this application, the starting point, data point, and prediction point all refer to the detection data obtained in each periodic inspection of the detection item corresponding to the inspection object.

[0133] In the state transition probability prediction model of the present application, if the reference value of the state transition probability prediction model is the above-mentioned state transition reference inspection data, then this model can be used to predict the state transition probability of the prediction point; if the reference value of the state transition probability prediction model is replaced with the actual historical inspection data corresponding to the inspection item, then it is to train the state transition probability prediction model. The specific training process is to estimate or adjust the values of the parameters in the model by bringing the historical actual historical inspection data into the model, so that the model can better reflect the actual situation. Specifically, the actual historical inspection data is used to optimize the detection state transition probability and the median value matrix W of the state space of each detection state in the state transition probability matrix P in the model, T so as to improve the accuracy of prediction.

[0134] S105. Use the state transition probability prediction model to predict the current inspection data in the inspection item, and obtain the probability of the current inspection data transitioning to the next state.

[0135] Obtain the current inspection data in the inspection item. When the number of inspections corresponding to the current inspection data in the inspection item is known, calculate the difference between the number of inspections corresponding to the current inspection data and the state transition starting point in the state transition probability prediction model, and bring the obtained difference into the state transition probability prediction model for prediction, to obtain the probability of the current inspection data transitioning to the next state.

[0136] Exemplarily, the current inspection item is to inspect the wear amount of the railway track. The current plan for inspecting the wear amount of the railway track is to inspect once a month, so the interval between each data point is 1 month. Based on the technical solution of the present application, when formulating the actual inspection plan, use the state transition probability prediction model to predict the data of the wear condition of the railway track in the current inspection item, and it will be judged whether the wear amount of the railway reaches the transition state, and then obtain the probability of the state transition of the inspection data next month. If the state of a certain railway is likely to remain in good condition next month, then the inspection work for this railway will not be arranged. Otherwise, it will be arranged. In this way, the operation and maintenance of the wear degree of the railway track is completed.

[0137] The technical solution in the embodiment of the present application has at least the following technical effects or advantages:

[0138] In the technical solution provided by this application, the historical inspection data of the corresponding inspection items of the operation and maintenance equipment is preprocessed to obtain the historical inspection data related to the number of inspections. Combined with the state transition reference inspection data of the whitening differential equation detection item, and then combined with the state transition probability matrix and the state median matrix, a state transition probability prediction model is constructed. The state transition probability prediction model is used to predict the current inspection data in the detection item, and the probability of the current inspection data transferring to the next state is obtained. In the technical solution of this application, after the operation and maintenance equipment completes an inspection, the probability of the inspection data transferring to the next state can be obtained. Therefore, when constructing the inspection plan, there is no need to set the inspection plan according to a fixed plan. The inspection plan can be transformed from a fixed shift inspection plan to an inspection based on the state. Furthermore, the density of the inspection work can be effectively reduced while ensuring that potential safety hazards can be detected in a timely manner during the inspection work.

[0139] Figure 2 As shown in the figure, it is a schematic structural diagram of a prediction system for inspection data of an operation and maintenance equipment in an embodiment of this application. The system includes:

[0140] An acquisition module, configured to acquire the first historical inspection data of the corresponding inspection items of the operation and maintenance equipment;

[0141] A processing module, configured to preprocess the first historical inspection data of the inspection items to obtain the second historical inspection data related to the number of inspections;

[0142] A determination module, configured to determine the state transition reference inspection data of the inspection items based on the second historical inspection data and the inspection data state transition model;

[0143] A construction module, configured to construct a state transition probability prediction model based on the second historical inspection data, the state transition reference inspection data, and the state transition probability matrix;

[0144] A prediction module, configured to use the state transition probability prediction model to predict the current inspection data in the inspection items, and obtain the probability of the current inspection data transferring to the next state.

[0145] Furthermore, the processing module is further configured to:

[0146] Normalize the discrete data obtained from each active detection in the first historical inspection data; and / or

[0147] Divide the continuous data in the first historical inspection data into intervals, and convert the data within the intervals using a conversion coefficient; and / or

[0148] Record the duration of each state corresponding to each state switch for the state data in the first historical inspection data.

[0149] Further, the determination module is further configured to:

[0150] Accumulate the first data point in the second historical inspection data to the current data point to obtain the third historical inspection data;

[0151] Based on the second historical inspection data and the third historical inspection data, construct an inspection data state transition model;

[0152] Based on the second historical inspection data and the unbiased parameters of the inspection data state transition model, determine the state transition reference inspection data of the detection item.

[0153] Further, the construction module is further configured to:

[0154] Establish a deviation value sequence between the second historical inspection data and the state transition reference inspection data;

[0155] Based on the deviation value sequence and different thresholds or ranges, divide the detection items into several detection states;

[0156] Based on several detection states, establish a state transition probability matrix;

[0157] Based on the median of the state space of each detection state in several detection states, establish a state median matrix;

[0158] Based on the state transition reference inspection data, the state transition probability matrix, and the state median matrix, construct a state transition probability prediction model.

[0159] It should be noted that for the sake of convenience of description, Figure 2 Exemplarily, only the main modules of the prediction system structure of the operation and maintenance equipment inspection data are shown. In actual applications, the system may also include modules or components not shown in the figure; the system is not limited to the above module structure and may also be other module structures for implementing the above method embodiments.

[0160] Figure 3 The following is a schematic structural diagram of an electronic device provided in an embodiment of the present application. As shown in the figure, the electronic device includes: a processor and a memory;

[0161] Wherein, the processor is configured to read and execute the programs and instructions stored in the memory, so that the electronic device executes the above method embodiments.

[0162] It should be noted that for the sake of convenience of description, Figure 3 Exemplarily, only the main components of the electronic device are shown. In actual applications, the electronic device may also include components or components not shown in the figure.

[0163] The present application also provides a computer-readable storage medium storing a program or instructions. When the computer reads and executes the program or instructions, the computer is caused to execute the method embodiments described above.

[0164] Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A prediction method for operation and maintenance equipment inspection data, characterized in that, Including: Obtain the first historical inspection data of the inspection items corresponding to the operation and maintenance equipment; Preprocess the first historical inspection data of the inspection items to obtain the second historical inspection data related to the number of inspections; Based on the second historical inspection data and the inspection data state transition model, determine the state transition reference inspection data of the inspection items; Based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix, construct a state transition probability prediction model; Use the state transition probability prediction model to predict the current inspection data in the inspection items, and obtain the probability of the current inspection data transitioning to the next state.

2. The prediction method for operation and maintenance equipment inspection data according to claim 1, wherein The preprocessing of the first historical inspection data of the inspection items includes: Normalize the discrete data obtained by each active inspection in the first historical inspection data; and / or Divide the continuous data in the first historical inspection data into intervals, and convert the data within the intervals using a conversion coefficient; and / or For the state data in the first historical inspection data, record the duration corresponding to each state each time the state is switched.

3. The prediction method for operation and maintenance equipment inspection data according to claim 1, characterized in that, The determining the state transition reference inspection data of the inspection items based on the second historical inspection data and the inspection data state transition model includes: Accumulate the first data point in the second historical inspection data to the current data point to obtain the third historical inspection data; Based on the second historical inspection data and the third historical inspection data, construct an inspection data state transition model; Based on the unbiased parameters of the second historical inspection data and the inspection data state transition model, determine the state transition reference inspection data of the inspection items.

4. The prediction method for operation and maintenance equipment inspection data according to any one of claims 1-3, characterized in that, The constructing a state transition probability prediction model based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix includes: Establish a deviation value sequence between the second historical inspection data and the state transition reference inspection data; Based on the deviation value sequence and different thresholds or ranges, divide the inspection items into several inspection states; Based on several inspection states, establish a state transition probability matrix; Based on the median of the state space of each inspection state in several inspection states, establish a state median matrix; Based on the state transition reference inspection data, the state transition probability matrix and the state median matrix, construct a state transition probability prediction model.

5. A prediction system for operation and maintenance equipment inspection data, characterized in that, Including: An acquisition module for acquiring the first historical inspection data of the inspection items corresponding to the operation and maintenance equipment; A processing module for preprocessing the first historical inspection data of the inspection items to obtain the second historical inspection data related to the number of inspections; A determination module for determining the state transition reference inspection data of the inspection items based on the second historical inspection data and the inspection data state transition model; A construction module for constructing a state transition probability prediction model based on the second historical inspection data, the state transition reference inspection data and the state transition probability matrix; A prediction module for using the state transition probability prediction model to predict the current inspection data in the inspection items, and obtaining the probability of the current inspection data transitioning to the next state.

6. The prediction system for operation and maintenance equipment inspection data according to claim 5, characterized in that, The processing module is further configured to: Normalize the discrete data obtained by each active inspection in the first historical inspection data; and / or Divide the continuous data in the first historical inspection data into intervals, and convert the data within the intervals using a conversion coefficient; and / or For the status data in the first historical inspection data, record the duration corresponding to each status each time the status is switched.

7. The prediction system for operation and maintenance equipment inspection data according to claim 5, characterized in that, The determining module is further configured to: Accumulate the first data point in the second historical inspection data to the current data point to obtain the third historical inspection data; Based on the second historical inspection data and the third historical inspection data, construct an inspection data state transition model; Based on the second historical inspection data and the unbiased parameters of the inspection data state transition model, determine the state transition reference inspection data of the inspection item.

8. The prediction system for operation and maintenance equipment inspection data according to any one of claims 5-7, characterized in that The constructing module is further configured to: Establish a deviation value sequence between the second historical inspection data and the state transition reference inspection data; Based on the deviation value sequence and different thresholds or ranges, divide the inspection item into several inspection states; Based on several inspection states, establish a state transition probability matrix; Based on the median of the state space of each inspection state among several inspection states, establish a state median matrix; Based on the state transition reference inspection data, the state transition probability matrix and the state median matrix, construct a state transition probability prediction model.

9. An electronic device, characterized in that, Comprising: A processor and a memory; The processor is coupled to the memory; Wherein, the processor is configured to read and execute the program or instruction stored in the memory, so that the device executes the method according to any one of claims 1-4.

10. A computer-readable storage medium, characterized in that, A computer program is stored, and when the program is executed by the processor, the method according to any one of claims 1-4 is implemented.

Images