A method and device for monitoring the temperature distribution performance of downhole circuit boards

Abstract

This invention discloses a method and device for monitoring the temperature distribution performance of downhole circuit boards. The method includes: first, constructing a sparse dictionary using a dictionary learning algorithm and setting an observation matrix; then, deploying a sparse temperature acquisition sensor array; building an intelligent sensing model to collect and transmit real-time data of the sparse signal of the downhole circuit board temperature distribution and storing it in the cloud as historical data of the sparse signal; generating standard operating mode data features; generating a health status curve; and finally, using a binary classification method to analyze the health status curve to achieve early warning of abnormal heat sources on the downhole circuit board. This invention uses intelligent sensing technology and cloud technology to achieve block-based sparse acquisition and cloud computing of downhole circuit board performance characteristics, reducing the storage and computational load of the device. It employs a signal recovery and reconstruction algorithm to comprehensively characterize the temperature performance image of the downhole circuit board and to observe and warn of the heat source distribution of the downhole circuit in real time.

Description

Technical Field

[0001] This invention relates to the field of petroleum exploration technology, and in particular to a method and device for monitoring the temperature distribution performance of downhole circuit boards. Background Technology

[0002] During downhole operations in oil exploration and extraction, the temperature changes of electronic components in the drilling circuit need to be monitored at all times. Timely monitoring of the real-time temperature distribution of the circuit, timely warning of abnormal temperature heat sources, and taking reasonable and effective measures can prevent excessive damage to electronic components and damage to the overall circuit. Reasonable recording and effective monitoring of the temperature changes of the overall circuit plays a crucial role in the stable operation of the circuit board inside the downhole drill pipe.

[0003] The core of heat source monitoring for electronic components in downhole drilling circuits is data analysis. On the one hand, in order to detect electronic component damage in a timely manner and improve heat source detection efficiency, it is necessary to perform temperature data analysis and heat source distribution image display on the downhole circuit board in a timely manner. However, the dense and constantly changing temperature information means that the sensing module needs to store a large amount of data, which makes it difficult for subsequent temperature performance analysis and processing. Moreover, the layout of the temperature acquisition circuit is not conducive to the design of the downhole circuit board. On the other hand, without heat source distribution images, abnormal heat sources cannot be detected in time, and downhole circuit board operations cannot be warned in time, which can easily damage the downhole circuit board operating inside the drill string, causing drill pipe failure and affecting drilling efficiency and downhole operation time.

[0004] The existing downhole pressure and temperature continuous monitoring device and method, patent number (CN105089640A), provides a downhole pressure and temperature continuous monitoring system and method, including a downhole pressure and temperature measurement and storage device and a wireless readback device. The downhole pressure and temperature measurement and storage device includes a measurement unit, a measurement storage unit, a measurement battery, and a wireless communication unit. The measurement unit measures temperature and pressure values. The wireless readback device includes a communication antenna, a readback storage unit, and a readback battery. The communication antenna communicates with the wireless communication unit, enabling the readback storage unit to record data from the measurement storage unit. This invention applies wireless data communication technology downhole to gas wells and combines it with a storage-type monitoring device to achieve continuous monitoring of gas wells without removing the production tubing string, using cables, or affecting gas well production. The data is then read back to the surface via the wireless readback device, providing a basis for operational measures and production capacity prediction. In actual use, downhole operations are characterized by long cycles and complex environments. The data transmission rate from downhole circuits to the surface is limited, so it is often impossible to locate the distribution of heat sources in a timely and accurate manner, or to provide timely and accurate early warnings of abnormal heat source locations and times. Summary of the Invention

[0005] This invention provides a method and apparatus for monitoring the temperature distribution performance of downhole circuit boards, which solves the problem in the prior art where the data transmission upload rate is limited, thus making it difficult to locate the distribution of heat sources in a timely and accurate manner, and to provide timely and accurate early warning of abnormal heat source locations and times.

[0006] On one hand, embodiments of the present invention provide a method for monitoring the temperature distribution performance of a downhole circuit board, including:

[0007] A sparse dictionary is constructed using a dictionary learning algorithm, and an observation matrix is ​​set up. A sparse temperature acquisition sensor array is then arranged according to the observation matrix.

[0008] An intelligent sensing model is built using the observation matrix. The intelligent sensing model is used to collect and transmit real-time data of the sparse signal of the temperature distribution of the downhole circuit board and store it in the cloud as historical data of the sparse signal.

[0009] Standard working mode data features are generated from the historical data;

[0010] A health status curve is generated by comparing the deviations between the characteristics of the real-time data and the characteristics of the standard working mode data using a swarm analysis algorithm.

[0011] A binary classification method is used to analyze the health status curve to achieve early warning of abnormal heat sources on downhole circuit boards.

[0012] One possible implementation also includes:

[0013] The temperature distribution data of the downhole circuit board in real time is recovered by using sparsely acquired temperature signals through signal reconstruction algorithms and sparse dictionaries, and a heat source distribution image is generated on the host computer.

[0014] Early warnings are issued for abnormal heat source locations based on the health status curve and the heat source distribution image.

[0015] In one possible implementation, the sparse dictionary is established by training historical temperature distribution data of downhole circuit boards using a dictionary learning algorithm to obtain temperature distribution performance characteristics, and then establishing the dictionary based on the temperature distribution performance characteristics.

[0016] In one possible implementation, the observation matrix is ​​established based on the characteristic relationships between atoms in the sparse dictionary.

[0017] On the other hand, embodiments of the present invention provide a downhole circuit board temperature distribution performance monitoring device, comprising:

[0018] Sensor module: Used to construct a sparse dictionary and set up an observation matrix using a dictionary learning algorithm, and to arrange a sparse temperature acquisition sensor array according to the observation matrix.

[0019] Sensing module: Used to build an intelligent sensing model through the observation matrix, collect and transmit real-time data of sparse signals of temperature distribution on downhole circuit boards through the intelligent sensing model and store it in the cloud as historical data of the sparse signals.

[0020] Analysis module: used to generate standard working mode data features from the historical data; and to generate a health status curve by comparing the deviation between the features of the real-time data and the features of the standard working mode data using a swarm analysis algorithm.

[0021] Early warning module: Used to analyze health status curves using a binary classification method to provide early warning of abnormal heat sources on downhole circuit boards.

[0022] In one possible implementation, the early warning module further includes a location early warning unit and a time early warning unit.

[0023] In one possible implementation, the location warning unit is used to recover the real-time acquired sparse signal data using the sparse dictionary and signal reconstruction algorithm, generate a heat source distribution image on the host computer, and issue warnings for abnormal heat source locations based on the health status curve and the heat source distribution image. The time warning unit determines the corresponding warning time based on the abnormal heat source location.

[0024] In one possible implementation, the sensor module and perception module are located locally, while the analysis module and early warning module are located in the cloud. The analysis module and early warning module transmit the analysis and early warning results to the host computer.

[0025] The method and apparatus for monitoring the temperature distribution performance of downhole circuit boards according to the present invention have the following advantages:

[0026] (1) Intelligent sensing technology and cloud technology are used to realize the sparse collection of performance characteristics of downhole circuit boards in blocks and cloud computing, which reduces the storage and computing power of the device.

[0027] (2) The temperature sparse acquisition sensor array is arranged according to the set observation matrix, which simplifies the design of the temperature acquisition circuit of the downhole circuit board.

[0028] (3) A signal recovery and reconstruction algorithm is used to fully characterize the temperature performance image of the downhole circuit board with fewer sparse features, and to observe the heat source distribution and abnormal heat source distribution location of the downhole circuit.

[0029] (4) A health status curve is established by using a group analysis algorithm, and the deviation of the health status curve is analyzed by a binary classification method to realize real-time early warning of abnormal heat sources. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a schematic diagram of a downhole circuit board temperature distribution performance monitoring device provided in an embodiment of the present invention.

[0032] Figure 2 This is a schematic diagram of a method for monitoring the temperature distribution performance of a downhole circuit board, provided in an embodiment of the present invention.

[0033] Figure 3 This is a schematic diagram of the sensor block acquisition downhole circuit board temperature distribution structure provided in an embodiment of the present invention.

[0034] Figure 4 This is a schematic diagram of the overall process of the intelligent sensing model provided in an embodiment of the present invention.

[0035] Figure 5 This is a schematic diagram of the atomic number adaptive KSVD algorithm provided in an embodiment of the present invention.

[0036] Figure 6 This is a schematic diagram of the working mode of the analysis module provided in an embodiment of the present invention.

[0037] Figure 7 This is a schematic diagram of the signal recovery and reconstruction algorithm provided in an embodiment of the present invention.

[0038] Figure 8 This is a schematic diagram illustrating the effect of establishing a health status curve by the host computer in an embodiment of the present invention.

[0039] Figure 9 This is a flowchart illustrating the operation of the upper-position display interface provided in an embodiment of the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] Figure 1This is a schematic diagram of a downhole circuit board temperature distribution performance monitoring device provided in an embodiment of the present invention; the present invention provides a downhole circuit board temperature distribution performance monitoring device, comprising:

[0042] Sensor module: used to construct a sparse dictionary and set up an observation matrix through a dictionary learning algorithm, and to arrange a sparse temperature acquisition sensor array according to the observation matrix;

[0043] Sensing module: used to build an intelligent sensing model through the observation matrix, collect and transmit real-time data of sparse signals of temperature distribution on downhole circuit boards through the intelligent sensing model and store them in the cloud as historical data of the sparse signals;

[0044] Analysis module: used to generate standard operating mode data features from the historical data; and to generate a health status curve by comparing the deviation between the features of the real-time data and the features of the standard operating mode data using a swarm analysis algorithm.

[0045] Early warning module: Used to analyze health status curves using a binary classification method to provide early warning of abnormal heat sources on downhole circuit boards.

[0046] In one possible embodiment, the sensor module and perception module are located locally, while the analysis module and early warning module are located in the cloud and transmit the analysis and early warning results to the host computer.

[0047] This invention provides a method for monitoring the temperature distribution performance of downhole circuit boards, such as... Figure 2 As shown, it includes:

[0048] A sparse dictionary is constructed using a dictionary learning algorithm, and an observation matrix is ​​set up. A sparse temperature acquisition sensor array is then arranged according to the observation matrix.

[0049] An intelligent sensing model is built using the observation matrix. The intelligent sensing model is used to collect and transmit real-time data of the sparse signal of the temperature distribution of the downhole circuit board and store it in the cloud to generate historical data of the sparse signal.

[0050] Standard working mode data features are generated from the historical data;

[0051] A health status curve is generated by comparing the deviations between the characteristics of the real-time data and the characteristics of the standard working mode data using a swarm analysis algorithm.

[0052] A binary classification method is used to analyze the health status curve to achieve early warning of abnormal heat sources on downhole circuit boards.

[0053] For example, such as Figure 3As shown, the circuit board for downhole operations is divided into 12*24 areas, each of which can serve as a point for downhole temperature acquisition. The temperature signal acquisition unit corresponds to each of the 288 areas on the downhole circuit board. Twenty sensors are sparsely arranged using an observation matrix to collect signals from these 20 areas and store them in a memory. The memory unit contains a unified numbered storage of the data randomly collected by the 20 temperature sensors.

[0054] The sensing module connects to a sparse temperature acquisition sensor array to collect and store sparse signals, which are then transmitted over a network to a remote server, i.e., a host computer. In the host computer, a swarm analysis algorithm is used to compare the deviation between the characteristics of the real-time collected sparse signals and the characteristics of the standard working mode data, and a health status curve is generated. The early warning module uses a binary classification method to analyze the health status curve and realize early warning of the temperature of the downhole circuit board.

[0055] The health status curve is generated by comparing the sparse signal features acquired in real time with the standard operating mode data features, and is updated in real time. The standard operating mode data features are summarized and generalized from the previous sparse signals using a group analysis algorithm, and are the standard data features for the normal operation of the downhole circuit board. By comparing the standard operating mode data with the real-time health status curve, and using a binary classification method to set the deviation alarm threshold of the health status curve, real-time monitoring and early warning of the downhole circuit board's operating temperature can be achieved.

[0056] In one possible embodiment, the observation matrix is ​​established based on the characteristic relationships between atoms in the sparse dictionary.

[0057] For example, such as Figure 4 As shown, the observation matrix is ​​determined based on the dictionary atomic feature distribution. The projection observation process is actually the process of mapping the electrical signal from a high-dimensional space to a low-dimensional space through the measurement matrix, that is:

[0058] T = ΦT out

[0059] In the formula: Φ is the measurement matrix, and T is the N-dimensional temperature signal T. out The M-dimensional observation values ​​obtained after the projection observation process, M < <N。

[0060] To recover the original signal T from a small number of low-dimensional observations T out The perception matrix must satisfy the constraint of isometry. The correlation between the measurement matrix Φ and the sparse dictionary Ψ can be used as an equivalent condition for RIP decision. The correlation between the two matrices Φ and Ψ can be calculated using the following formula:

[0061]

[0062] The greater the correlation between Φ and Ψ, the larger the correlation coefficient μ(Φ,Ψ), and vice versa. Clearly, the smaller the correlation coefficient μ(Φ,Ψ), the less the original signal T carried by the compressed measurement T. out The more information, the better the reconstruction of the original signal T. out The higher the accuracy.

[0063] In one possible embodiment, the sparse dictionary is established by training historical temperature distribution data of downhole circuit boards using a dictionary learning algorithm to obtain temperature distribution performance characteristics, and then establishing the sparse dictionary based on the temperature distribution performance characteristics.

[0064] For example, the sparse dictionary is established by training the first pair of historical temperature distribution data of downhole circuit boards using the atom-adaptive KSVD algorithm to obtain temperature distribution performance characteristics, and then establishing the sparse dictionary. The method is as follows:

[0065] First, the historical signals of the sparse signal collected by the intelligent sensing model are used to train the temperature distribution performance characteristics using an adaptive KSVD algorithm. The intelligent sensing model is T = T out =ΦΨX=AX, where, T out —Original temperature signal, N×1 matrix; Ψ —Sparse dictionary, N×L matrix; X —Projection coefficients of the original signal onto the dictionary, L×1 matrix, X=[x1,x2,x3,...,x L ] T The number of non-zero elements L in X does not exceed K (where K << N). Φ is the observation matrix, an M×N matrix, and A is the perception matrix, an M×L matrix.

[0066] Then, as Figure 5 As shown, the sparse dictionary construction steps for obtaining the temperature performance distribution image features of the downhole circuit board are as follows:

[0067] Step 1: Initialize the dictionary Ψ as a normalized random matrix, and set the number of dictionary atoms l and the number of iterations k based on experience;

[0068] Step 2: Using the OMP reconstruction algorithm, obtain the sparse coefficient matrix X from the historical dataset Y and the dictionary Ψ;

[0069] Step 3: Calculate the reconstruction error matrix Ω using the formula. j0 ;

[0070]

[0071] Step 4: Determine if the reconstruction error is less than the threshold. If it is, discard the atoms; otherwise, proceed to the next step.

[0072] Step 5: Update each atom in the dictionary Ψ sequentially, that is, update Ω in Step 3. j0 Perform SVD decomposition on (j0 = 1, 2, ...) to obtain the characterization of Ω. j0 The largest eigenvector is used as the new atom to replace the original ψ. j0 ;

[0073] Step 6: Repeat steps 2-4 until the set number of iterations is reached, then output the optimal dictionary Ψ.

[0074] As the dictionary is continuously projected, the number of dictionary atoms l increases, the dictionary's expressive power becomes stronger, the number of features increases, and its ability to fully characterize signals becomes more complete, making the representativeness of the temperature performance distribution image features of the downhole circuit board more obvious.

[0075] The intelligent sensing model trains features for each sensor using the KSVD algorithm on data collected from different blocks.

[0076] For example, such as Figure 6 As shown, the analysis module includes two modes: "offline learning" and "online working". The "offline learning" mode includes using the KSVD algorithm to train historical data to obtain a sparse dictionary and using a clustering algorithm to train atomic features to obtain the device's standard working mode. The "online working" mode includes acquiring real-time collected features to establish a health status curve.

[0077] The KSVD algorithm used in the offline learning mode requires first reconstructing the collected sparse data using a signal recovery and reconstruction algorithm to create historical data. This historical data is stored on a host computer and used for feature training. The data storage includes commands sent by the host computer to read temperature signals via a network connection to the sensor interface after the downhole operation concludes.

[0078] Furthermore, the online learning mode acquires sparse sampled values ​​in real time during downhole operations and uses a method of establishing health status curves to analyze and provide early warnings regarding the temperature performance characteristics distribution of the downhole circuit board's operating status. The health status curves and quality reports are generated by the host computer and then displayed on the host computer's temperature image performance display interface, while the host computer's management system synchronously stores them.

[0079] Then, the health status curve is analyzed, and a correlation fitting maximizing similarity is performed between the fault operation mode deviation and the actual downhole circuit board temperature distribution. An alarm threshold is set to achieve automatic identification of abnormal heat sources in the downhole operating device. The fault operation mode deviation represents the deviation between the current operating condition of the device and the standard operating mode obtained from cluster analysis. The larger the operation mode deviation, the greater the difference between the device's operating condition and the standard operating mode. Under normal operating conditions, the operation mode deviation should be less than a certain set threshold. When the operation mode deviation exceeds the set threshold, it is determined that there is an abnormal heat source in the downhole operating device, indicating an abnormal operating mode, and an automatic alarm is triggered.

[0080] In one possible embodiment, the early warning module further includes a location early warning unit and a time early warning unit. The location early warning unit is used to recover the real-time acquired sparse signal data using a sparse dictionary and signal reconstruction algorithm, generate a heat source distribution image on the host computer, and issue early warnings for abnormal heat source locations based on the standard operating mode data and the heat source distribution image. The time early warning unit determines the corresponding time based on the abnormal heat source location.

[0081] For example, the host computer can generate a complete heat source distribution image by restoring historical data, and establish a dynamic graph of the heat source distribution image changes over a long period of time by establishing the time relationship between operation and sampling. The interface of the host computer displays the heat source distribution of the temperature image and provides timely warnings for high-temperature heat sources. The host computer is a computer.

[0082] like Figure 7 As shown, the signal recovery and reconstruction algorithm steps are as follows:

[0083] Step 1: Input the sparse sampling sequence, observation matrix, reconstruction allowable error, and sparse dictionary;

[0084] Step 2: Calculate the perception matrix and the residual vector, i.e.:

[0085]

[0086] In the formula: Let A be the transpose of the atom in the j-th column of the perception matrix A.

[0087] Step 3: The atom with the highest correlation coefficient is the atom that best matches the signal. The index value λ corresponding to the highest correlation coefficient is then updated in the support set, i.e., s. i =s i-1 ∪{λ}, update the column vector set B i =[B i-1 A λ ].

[0088] Step 4: Apply the least squares method to the column atom set Bi To approximate the coefficients, that is:

[0089] X # = (B i T B i ) -1 B i T r

[0090] The residual vector is then updated using the following formula:

[0091] r=TB i -B i X #

[0092] Step 5: Repeat steps 2-4 until the magnitude of the residual vector r is less than the reconstruction allowable error ε, thus obtaining the sparse coefficients X. # Finally, the reconstructed signal T is calculated using the formula. out .

[0093] The cluster analysis algorithm includes clustering the performance distribution features of temperature images obtained by KSVD training, so that each cluster represents the standard temperature pattern at the corresponding location in the image.

[0094] For example, such as Figure 8 As shown, sparse sampling data of state parameters obtained from historical heat source distribution images are projected onto a pre-built sparse dictionary. The K-means clustering algorithm is used to classify the projection coefficients into k classes, so that each class represents a standard operating mode of the device.

[0095] Step 1: Cluster the data using k points in the space as initial centers, calculate the distance between historical data and cluster centers, and select the cluster number closest to the data sample as the sample's operating mode.

[0096]

[0097]

[0098] Where m j p is the cluster index that is closest to the cluster center in historical data. j y is the coordinate of the i-th cluster center; C is the set of all cluster centers; y is one historical data sample; d min It is the shortest distance from historical data to the cluster center.

[0099] Step 2: Use the mean method to update the cluster center value of the class. The new cluster center is the cluster center of each cluster, which is the mean of the cluster members. For all k cluster centers, if the cluster centers remain unchanged after the iterative update, the iteration ends. Each cluster center is regarded as a standard operating mode.

[0100] In one possible embodiment, the method further includes: recovering the sparse signal acquired in real time using a sparse dictionary and a signal reconstruction algorithm, and generating a heat source distribution image on a host computer;

[0101] Early warnings are issued for abnormal heat source locations based on the health status curve and the heat source distribution image.

[0102] For example, the temperature sparse features of the downhole circuit board acquired by the observation matrix are recovered using a signal recovery and reconstruction algorithm. The analysis module generates a heat source distribution image, which fully characterizes the temperature performance of the downhole circuit board with fewer sparse features. The heat source distribution of the downhole circuit is observed in real time, and abnormal heat source locations are monitored for early warning.

[0103] In one possible embodiment, such as Figure 9 As shown, the usage process of the monitoring device in this invention is as follows:

[0104] Open the temperature image display page on the host computer. Users can manually enter information such as the downhole operation location, operators, and downhole circuit boards in the information entry interface. The downhole circuit board information includes: the circuit characteristics of the downhole circuit board, the temperature characteristics of sensitive components, etc. The information entry interface automatically captures the operation time and signal acquisition time transmitted by the downhole acquisition device. Click the start option to begin the current downhole circuit board heat source distribution analysis.

[0105] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0106] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for monitoring the temperature distribution performance of a downhole circuit board, characterized in that, include: A sparse dictionary is constructed using a dictionary learning algorithm, and an observation matrix is ​​set up. A temperature sparse acquisition sensor array is then arranged according to the observation matrix. An intelligent sensing model is built using the observation matrix. The intelligent sensing model is used to collect and transmit real-time data of the sparse signal of the temperature distribution of the downhole circuit board and store it in the cloud as historical data of the sparse signal. Standard working mode data features are generated from the historical data; A health status curve is generated by comparing the deviations between the characteristics of the real-time data and the characteristics of the standard working mode data using a swarm analysis algorithm. A binary classification method is used to analyze the health status curve to achieve early warning of abnormal heat sources on downhole circuit boards; The temperature distribution data of the downhole circuit board in real time is recovered by using the sparsely acquired temperature signal through a signal reconstruction algorithm and the sparse dictionary, and a heat source distribution image is generated on the host computer. Early warnings are issued for abnormal heat source locations based on the health status curve and the heat source distribution image.

2. The method for monitoring the temperature distribution performance of a downhole circuit board according to claim 1, characterized in that, The sparse dictionary is established by training historical temperature distribution data of downhole circuit boards using a dictionary learning algorithm to obtain temperature distribution performance characteristics, and then establishing these temperature distribution performance characteristics.

3. The method for monitoring the temperature distribution performance of a downhole circuit board according to claim 1, characterized in that, The observation matrix is ​​established based on the characteristic relationships between the atoms in the sparse dictionary.

4. A device for monitoring the temperature distribution performance of a downhole circuit board, characterized in that, include: Sensor module: used to construct a sparse dictionary and set up an observation matrix through a dictionary learning algorithm, and to arrange a sparse temperature acquisition sensor array according to the observation matrix; Sensing module: used to build an intelligent sensing model through the observation matrix, collect and transmit real-time data of sparse signals of temperature distribution on downhole circuit boards through the intelligent sensing model and store them in the cloud to generate historical data of the sparse signals; Analysis module: used to generate standard working mode data features from the historical data; A health status curve is generated by comparing the deviations between the characteristics of the real-time data and the characteristics of the standard working mode data using a swarm analysis algorithm. Early warning module: used to analyze the health status curve using a binary classification method to achieve early warning of abnormal heat sources on downhole circuit boards; the early warning module also includes a location early warning unit and a time early warning unit; the location early warning unit is used to recover the temperature distribution data of the downhole circuit board collected in real time through the sparse dictionary and signal reconstruction algorithm, generate a heat source distribution image on the host computer, and issue early warnings for abnormal heat source locations based on the health status curve and the heat source distribution image; the time early warning unit determines the corresponding early warning time based on the abnormal heat source location.

5. The downhole circuit board temperature distribution performance monitoring device according to claim 4, characterized in that, The sensor module and perception module are located locally, while the analysis module and early warning module are located in the cloud. The analysis module and early warning module transmit the analysis and early warning results to the host computer.

Images