Data fusion transmission method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of data fusion transmission, in particular to a data fusion transmission method and device, electronic equipment and a storage medium. Cutting the plurality of data streams, and performing data superposition processing on a plurality of arrays obtained by cutting to obtain a preprocessing coefficient and a plurality of first arrays; adjusting a plurality of parameters of the first data fusion model pair by using a plurality of first arrays to obtain a second data fusion model pair; and finally, inputting the plurality of first arrays into a second coding model to obtain a plurality of second arrays, and transmitting the plurality of target model parameters, the plurality of preprocessing coefficients and the plurality of second arrays. According to the method, the data are subjected to superposition processing, the fusion model is constructed by utilizing the superposition ratio of the data, the data with the high superposition ratio is represented by the fusion model through short codes, the length of the fused data is shortened, the total data transmission amount obtained through fusion is smaller, and the data transmission efficiency is higher.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion transmission, and in particular to a data fusion transmission method, device, electronic equipment and storage medium. Background Art

[0002] The power inspection device is mainly used for the survey, maintenance and inspection of power lines. The power inspection device has the characteristics of multi-purpose and diverse working scenarios, which requires the power inspection device to collect a variety of data including images and various sensor data during operation.

[0003] At the same time, the power inspection device needs to follow the command of the rear platform during operation, and a large amount of collected data needs to be transmitted to the rear command platform. The transmission of a large amount of data limits the reliability and timeliness of data transmission.

[0004] One solution is to classify the transmitted data into multiple priorities and transmit the data according to the priority of the data. This solution does not substantially improve the efficiency of data transmission.

[0005] Based on this, it is necessary to develop and design a data fusion transmission method. Summary of the invention

[0006] The embodiments of the present invention provide a data fusion transmission method, device, electronic device and storage medium, which are used to solve the problem of low data transmission efficiency in the prior art.

[0007] In a first aspect, an embodiment of the present invention provides a data fusion transmission method, including: Get multiple data streams; Cutting the multiple data streams, performing data overlap processing on the multiple arrays obtained by cutting, and obtaining multiple preprocessing coefficients and multiple first arrays; Using the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; The multiple first arrays are input into the second encoding model to obtain multiple second arrays, and multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays are transmitted, wherein the target model parameters are parameters of the second decoding model.

[0008] In a possible implementation, the cutting of the multiple data streams, performing data overlap processing on the multiple arrays obtained by cutting, and obtaining multiple preprocessing coefficients and multiple first arrays include: Get the mapping interval; Cut the multiple data streams respectively according to the first data length to obtain multiple intermediate arrays; Counting the number of values ​​appearing in the plurality of intermediate arrays respectively to obtain a plurality of first statistical numbers, wherein each first statistical number corresponds to an intermediate array, and the first statistical number represents the number of values ​​appearing in the intermediate array; If the maximum value among the plurality of first statistics is greater than the mapping interval, the first data length is reduced, and the process jumps to the step of cutting the plurality of data streams according to the first data length to obtain a plurality of intermediate arrays; Otherwise, the plurality of intermediate arrays are adjusted according to the mapping interval and the lower limit of the mapping interval to obtain the plurality of preprocessing coefficients and the plurality of first arrays.

[0009] In a possible implementation manner, adjusting the multiple intermediate arrays according to the mapping interval and the lower limit of the mapping interval to obtain the preprocessing coefficients and the multiple first arrays includes: For each intermediate array, extract the maximum data and the minimum data as two preprocessing coefficients of the intermediate array, thereby obtaining a plurality of preprocessing coefficients; According to the first formula, the mapping interval and the lower limit of the mapping interval, the multiple intermediate arrays are adjusted to obtain the multiple first arrays, wherein the first formula:

[0010] In the formula, is the mapping interval, is the lower limit of the mapping interval, The middle array data, is the maximum data of the middle array, is the minimum data of the middle array, The first array data, is the rounding function.

[0011] In a possible implementation, the first data fusion model pair includes a first encoding model and a first decoding model, and the using the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain the second data fusion model pair includes: Obtaining a mapping interval, a first encoding model, and a first decoding model; Arrange the data appearing in the plurality of first arrays according to the magnitude of the frequencies of appearance in the plurality of first arrays to obtain a sample queue; Constructing a second data queue arranged in order of size and filling the mapping interval, wherein the data in the second data queue is an integer; constructing the second data queue as a label queue of the sample queue, wherein the label value of the sample queue data is negatively correlated with the occurrence frequency of the sample queue data in the plurality of first arrays; According to the sample queue and the label queue, multiple parameters of the first encoding model and multiple parameters of the first decoding model are adjusted, and the adjusted first encoding model and the adjusted first decoding model are used as a first data fusion model pair.

[0012] In a possible implementation manner, adjusting multiple parameters of the first encoding model and multiple parameters of the first decoding model according to the sample queue and the label queue, and using the adjusted first encoding model and the adjusted first decoding model as a first data fusion model pair, includes: Randomly inputting a plurality of data of the sample queue into the first coding model; If the deviation between the output of the first coding model and the target label is greater than a first threshold, multiple parameters of the first coding model are adjusted according to the deviation between the output of the first coding model and the target label, and the process jumps to the step of randomly inputting multiple data of the sample queue into the first coding model, wherein the target label is a label corresponding to the data input into the first coding model; Randomly inputting a plurality of labels of the label queue into the first decoding model; If the deviation between the output of the first decoding model and the target data is greater than a second threshold, multiple parameters of the first decoding model are adjusted according to the deviation between the output of the first decoding model and the target data, and the process jumps to the step of randomly inputting multiple labels of the label queue into the first decoding model, wherein the target data is data of the sample queue identified by the label input into the first decoding model; The adjusted first encoding model and the adjusted first decoding model are used as the second encoding model and the second decoding model respectively.

[0013] In a possible implementation, the first coding model is:

[0014] In the formula, is the output of the first coding model, For the Line The output of the intermediate function is as follows, is the node meta-function, is the first bias parameter, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, For the The output of the input node, For the The weight parameters of the input nodes, For the Input data, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column Weight parameters; The first decoding model is:

[0015] In the formula, is the output of the second model, For the Line The output of the intermediate function is as follows, is the second bias constant, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, is the output of the input node, is the weight parameter of the input node, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameter.

[0016] In one possible implementation, the node meta-function is:

[0017] In the formula, is the input quantity, is a natural constant.

[0018] In a second aspect, an embodiment of the present invention provides a data fusion transmission device, which is used to implement the data fusion transmission method described in the first aspect or any possible implementation of the first aspect, and the data fusion transmission device includes: A data stream acquisition module, used to acquire multiple data streams; A data adjustment module, used for cutting the multiple data streams, performing data overlap processing on the multiple arrays obtained by cutting, and obtaining preprocessing coefficients and multiple first arrays; A fusion model construction module, used to adjust multiple parameters of the first data fusion model pair using the multiple first arrays to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; as well as, A data transmission module is used to input the multiple first arrays into the second encoding model to obtain multiple second arrays, and to transmit multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays, wherein the target model parameters are the parameters of the second decoding model.

[0019] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the steps of the method described in the first aspect or any possible implementation method of the first aspect.

[0020] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect or any possible implementation of the first aspect are implemented.

[0021] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The data fusion transmission method implementation method of the present invention first obtains multiple data streams; then cuts the multiple data streams, performs data overlap processing on the multiple arrays obtained by cutting, obtains preprocessing coefficients and multiple first arrays; then uses the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; finally, the multiple first arrays are input into the second encoding model to obtain multiple second arrays, and multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays are transmitted, wherein the target model parameters are the parameters of the second decoding model. The implementation method of the present invention performs data overlap processing, and constructs a fusion model using the overlap of the data. The fusion model uses short coding to represent the data with high overlap, and the length of the fused data becomes shorter, the total amount of data transmission obtained by fusion is smaller, and the efficiency of data transmission is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, 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, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 is a flow chart of a data fusion transmission method provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a data stream cutting process provided by an embodiment of the present invention; Figure 3 It is a principle diagram of numerical statistics of array occurrences provided by an embodiment of the present invention; Figure 4 It is a functional block diagram of a data fusion transmission device provided by an embodiment of the present invention; Figure 5 It is a functional block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, and methods are omitted so as not to obscure the description of the present invention with unnecessary details.

[0025] In order to make the purpose, technical solutions and advantages of the present invention more clear, a specific implementation method will be described below in conjunction with the accompanying drawings.

[0026] The following is a detailed description of an embodiment of the present invention. This example is implemented based on the technical solution of the present invention, and provides a detailed implementation method and a specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0027] Figure 1 The present invention provides a flowchart of a data fusion transmission method according to an embodiment of the present invention.

[0028] like Figure 1 As shown, it shows a flow chart of the implementation of the data fusion transmission method provided by an embodiment of the present invention, which is described in detail as follows: In step 101, multiple data streams are obtained.

[0029] In step 102, the multiple data streams are cropped, and the multiple arrays obtained by the cropping are subjected to data overlap processing to obtain multiple preprocessing coefficients and multiple first arrays.

[0030] In some implementations, step 102 includes: obtaining a mapping interval; Cut the multiple data streams respectively according to the first data length to obtain multiple intermediate arrays; Counting the number of values ​​appearing in the plurality of intermediate arrays respectively to obtain a plurality of first statistical numbers, wherein each first statistical number corresponds to an intermediate array, and the first statistical number represents the number of values ​​appearing in the intermediate array; If the maximum value among the plurality of first statistics is greater than the mapping interval, the first data length is reduced, and the process jumps to the step of cutting the plurality of data streams according to the first data length to obtain a plurality of intermediate arrays; Otherwise, the plurality of intermediate arrays are adjusted according to the mapping interval and the lower limit of the mapping interval to obtain the plurality of preprocessing coefficients and the plurality of first arrays.

[0031] In some implementations, adjusting the plurality of intermediate arrays according to the mapping interval and the lower limit of the mapping interval to obtain the preprocessing coefficients and the plurality of first arrays includes: For each intermediate array, extract the maximum data and the minimum data as two preprocessing coefficients of the intermediate array, thereby obtaining a plurality of preprocessing coefficients; According to a first formula, the mapping interval and the lower limit of the mapping interval, the multiple intermediate arrays are adjusted to obtain the multiple first arrays, wherein the first formula:

[0032] In the formula, is the mapping interval, is the lower limit of the mapping interval, The middle array data, is the maximum data of the middle array, is the minimum data of the middle array, The first array data, is the rounding function.

[0033] Exemplarily, an embodiment of the present invention first obtains multiple data streams, which may correspond to multiple data collection points, such as temperature data, humidity data, location data, image data, and the like. Transmitting these data after fusion can reduce the total amount of data transmission and improve the efficiency of data transmission. In an embodiment of the present invention, multiple data streams are cut according to a preset length to obtain multiple arrays, and the multiple arrays are overlapped. Then, the frequency of occurrence of the overlapped data is counted, and the frequency of occurrence is used as a label of the data. A fusion model is established using these overlapped data. Finally, the overlapped data is used to generate fused data through the fusion model, and the fused data is transmitted. Since the fused data is based on the frequency of occurrence of the data, and the higher the frequency of occurrence of the data, the shorter the length of the fused data, the total amount of data transmission obtained by fusion is smaller and the efficiency of data transmission is higher.

[0034] To achieve the above purpose, the embodiment of the present invention firstly cuts and overlaps multiple data streams, obtains multiple intermediate arrays after cutting, and then counts the number of values ​​appearing in each intermediate array. If the number of values ​​appearing in these intermediate arrays is greater than the number of mapping intervals, then it is necessary to reduce the cutting length and re-cut; if it is less than the number of mapping intervals, then these intermediate arrays must be uniformly mapped into the mapping interval.

[0035] For example, after trimming the data stream, a total of N arrays from the 1st to the Nth are obtained, and the number of types of values ​​that appear in the Mth array is at most 1352 (for example, there are 2500 data in each array, and repeated values ​​are considered to be one value), and the range of the mapping interval is 1000 (0-999), which means that the trimming length is too long and needs to be shortened and then trimmed again.

[0036] In the aspect of performing the overlap processing, the implementation method of the present invention first extracts the maximum value and the minimum value in the array of each intermediate number as the preprocessing coefficient, thereby obtaining the preprocessing coefficient of twice the number of the numbers, and then performs the overlap processing on the intermediate array according to the first formula, and the first formula is:

[0037] In the formula, is the mapping interval, is the lower limit of the mapping interval, The middle array data, is the maximum data of the middle array, is the minimum data of the middle array, The first array data, is the rounding function.

[0038] Figure 2 The diagram shows a process of cutting a plurality of data streams 201 according to a preset data length 202 to obtain a plurality of intermediate arrays. Figure 3 The process of counting the number of values ​​302 that appear according to the intermediate array 301 is shown. In the figure, the same value is classified into one class, and the number of classes is the number of values ​​that appear in the intermediate array.

[0039] In step 103, multiple parameters of the first data fusion model pair are adjusted using the multiple first arrays to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model.

[0040] In some implementations, step 103 includes: Obtaining a mapping interval, a first encoding model, and a first decoding model; Arrange the data appearing in the plurality of first arrays according to the magnitude of the frequencies of appearance in the plurality of first arrays to obtain a sample queue; Constructing a second data queue arranged in order of size and filling the mapping interval, wherein the data in the second data queue is an integer; constructing the second data queue as a label queue of the sample queue, wherein the label value of the sample queue data is negatively correlated with the occurrence frequency of the sample queue data in the plurality of first arrays; According to the sample queue and the label queue, multiple parameters of the first encoding model and multiple parameters of the first decoding model are adjusted, and the adjusted first encoding model and the adjusted first decoding model are used as a first data fusion model pair.

[0041] In some implementations, the adjusting multiple parameters of the first encoding model and multiple parameters of the first decoding model according to the sample queue and the label queue, and using the adjusted first encoding model and the adjusted first decoding model as a first data fusion model pair, includes: Randomly inputting a plurality of data of the sample queue into the first coding model; If the deviation between the output of the first coding model and the target label is greater than a first threshold, multiple parameters of the first coding model are adjusted according to the deviation between the output of the first coding model and the target label, and the process jumps to the step of randomly inputting multiple data of the sample queue into the first coding model, wherein the target label is a label corresponding to the data input into the first coding model; Randomly inputting a plurality of labels of the label queue into the first decoding model; If the deviation between the output of the first decoding model and the target data is greater than a second threshold, multiple parameters of the first decoding model are adjusted according to the deviation between the output of the first decoding model and the target data, and the process jumps to the step of randomly inputting multiple labels of the label queue into the first decoding model, wherein the target data is data of the sample queue identified by the label input into the first decoding model; The adjusted first encoding model and the adjusted first decoding model are used as the second encoding model and the second decoding model respectively.

[0042] In some implementations, the first coding model is:

[0043] In the formula, is the output of the first encoding model, For the Line The output of the intermediate function is as follows, is the node meta-function, is the first bias parameter, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, For the The output of the input node, For the The weight parameters of the input nodes, For the Input data, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column Weight parameters; The first decoding model is:

[0044] In the formula, is the output of the second model, For the Line The output of the intermediate function is as follows, is the second bias constant, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, is the output of the input node, is the weight parameter of the input node, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameter.

[0045] In some implementations, the node meta-function is:

[0046] In the formula, is the input quantity, is a natural constant.

[0047] Exemplarily, in an embodiment of the present invention, the first array obtained in the above steps is used to adjust multiple parameters of the first data fusion model pair, so that the first data fusion model pair has the ability to encode the first array obtained in the above steps and the ability to decode according to the encoding. In fact, the first data fusion model pair includes two models, namely a first encoding model and a first set code model.

[0048] The first coding model is:

[0049] In the formula, is the output of the first coding model, For the Line The output of the intermediate function is as follows, is the node meta-function, is the first bias parameter, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, For the The output of the input node, For the The weight parameters of the input nodes, For the Input data, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column Weight parameters; The first decoding model is:

[0050] In the formula, is the output of the second model, For the Line The output of the intermediate function is as follows, is the second bias constant, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, is the output of the input node, is the weight parameter of the input node, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameter.

[0051] The node meta-function in the above formula is:

[0052] In the formula, is the input quantity, is a natural constant.

[0053] Before using multiple first arrays to adjust multiple parameters of the model queue, first arrange the data in the multiple first arrays according to the frequency of occurrence to obtain a sample queue, then fill it with integer corresponding number intervals to obtain a second data queue, and then construct the second data queue into a label queue of the sample queue. It should be emphasized here that the frequency of occurrence of the data in the sample queue is negatively correlated with the value of the label data.

[0054] For example, multiple first arrays are arranged in descending order according to the frequency of occurrence of numerical values, namely N1, N2, ... N999. At this time, a sample queue is obtained, and the data filling the mapping interval is arranged in ascending order to construct a second data queue, 1, 2, ... 999. A mapping is established between the two queues, then the corresponding position is its label.

[0055] After labeling the sample queue, data is extracted from the sample queue and input into the first coding model. According to the deviation between the output value of the coding model and the label of the input data, the parameters of the first coding model are adjusted (for example, using the gradient descent method). The above steps are repeatedly performed until the first coding model can accurately fit the relationship between the data and the label in the sample queue. At this time, the first coding model is used as the second coding model.

[0056] At the same time, data is extracted from the second data queue (label queue) and input into the first decoding model. The parameters of the first decoding model are adjusted according to the deviation between the output value of the decoding model and the data identified by the input data. The above steps are repeatedly performed until the first decoding model can accurately fit the relationship between the data and the label in the sample queue. At this time, the first decoding model is used as the second decoding model.

[0057] In step 104, the multiple first arrays are input into the second encoding model to obtain multiple second arrays, and multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays are transmitted, wherein the target model parameters are parameters of the second decoding model.

[0058] Exemplarily, after the encoding and decoding model is constructed, the data in multiple first arrays can be input into the second encoding model for encoding, and the obtained arrays are used as multiple second arrays to transmit the parameters of the second decoding model, multiple preprocessing coefficients and multiple second arrays.

[0059] After receiving the data at the receiving end, a decoding model is constructed according to the parameters of the second decoding model, and then multiple second arrays are input into the decoding model to obtain multiple third arrays. In fact, the third data is the restoration of the first array. The data is then de-coincided according to the first array and multiple preprocessing coefficients, and the data is restored.

[0060] The data fusion transmission method implementation method of the present invention first obtains multiple data streams; then cuts the multiple data streams, performs data overlap processing on the multiple arrays obtained by cutting, obtains preprocessing coefficients and multiple first arrays; then uses the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; finally, the multiple first arrays are input into the second encoding model to obtain multiple second arrays, and multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays are transmitted, wherein the target model parameters are the parameters of the second decoding model. The implementation method of the present invention performs data overlap processing, and constructs a fusion model using the overlap of the data. The fusion model uses short coding to represent the data with high overlap, and the length of the fused data becomes shorter, the total amount of data transmission obtained by fusion is smaller, and the efficiency of data transmission is higher.

[0061] It should be understood that the size of the serial numbers of the steps in the above implementation does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation method of the present invention.

[0062] The following is an implementation of the device of the present invention. For details not described in detail, reference may be made to the corresponding method implementation described above.

[0063] Figure 4 is a functional block diagram of a data fusion transmission device provided by an embodiment of the present invention, referring to Figure 4 The data fusion transmission device includes: a data stream acquisition module 401, a data adjustment module 402, a fusion model construction module 403 and a data transmission module 404, wherein: A data stream acquisition module 401 is used to acquire multiple data streams; A data adjustment module 402 is used to cut the multiple data streams, perform data overlap processing on the multiple arrays obtained by cutting, and obtain preprocessing coefficients and multiple first arrays; A fusion model construction module 403 is used to adjust multiple parameters of the first data fusion model pair using the multiple first arrays to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; The data transmission module 404 is used to input the multiple first arrays into the second encoding model to obtain multiple second arrays, and transmit multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays, wherein the target model parameters are the parameters of the second decoding model.

[0064] Figure 5 is a functional block diagram of an electronic device provided by an embodiment of the present invention. Figure 5 As shown, the electronic device 5 of this embodiment includes: a processor 500 and a memory 501, wherein the memory 501 stores a computer program 502 that can be run on the processor 500. When the processor 500 executes the computer program 502, the steps in the above-mentioned data fusion transmission method and the embodiment are implemented, for example Figure 1 Steps 101 to 104 are shown.

[0065] Exemplarily, the computer program 502 may be divided into one or more modules / units, and the one or more modules / units are stored in the memory 501 and executed by the processor 500 to implement the present invention.

[0066] The electronic device 5 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device 5 may include, but is not limited to, a processor 500 and a memory 501. Those skilled in the art will appreciate that Figure 5 It is only an example of the electronic device 5 and does not constitute a limitation of the electronic device 5. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 5 may also include input and output devices, network access devices, buses, etc.

[0067] The processor 500 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0068] The memory 501 may be an internal storage unit of the electronic device 5, such as a hard disk or memory of the electronic device 5. The memory 501 may also be an external storage device of the electronic device 5, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 5. Further, the memory 501 may also include both an internal storage unit of the electronic device 5 and an external storage device. The memory 501 is used to store the computer program 502 and other programs and data required by the electronic device 5. The memory 501 may also be used to temporarily store data that has been output or is to be output.

[0069] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the implementation method can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method implementation method, which will not be repeated here.

[0070] In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0071] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0072] In the embodiments provided by the present invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0073] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0074] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0075] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned implementation method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various methods and device implementation methods can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0076] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A data fusion transmission method, characterized in that: include: Get multiple data streams; Cutting the multiple data streams, performing data overlap processing on the multiple arrays obtained by cutting, and obtaining multiple preprocessing coefficients and multiple first arrays; Using the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; The multiple first arrays are input into the second encoding model to obtain multiple second arrays, and multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays are transmitted, wherein the target model parameters are parameters of the second decoding model.

2. The data fusion transmission method according to claim 1, characterized in that: The cutting of the multiple data streams and the data overlap processing of the multiple arrays obtained by cutting to obtain multiple preprocessing coefficients and multiple first arrays include: Get the mapping interval; Cut the multiple data streams respectively according to the first data length to obtain multiple intermediate arrays; Counting the number of values ​​appearing in the plurality of intermediate arrays respectively to obtain a plurality of first statistical numbers, wherein each first statistical number corresponds to an intermediate array, and the first statistical number represents the number of values ​​appearing in the intermediate array; If the maximum value among the plurality of first statistics is greater than the mapping interval, the first data length is reduced, and the process jumps to the step of cutting the plurality of data streams according to the first data length to obtain a plurality of intermediate arrays; Otherwise, the plurality of intermediate arrays are adjusted according to the mapping interval and the lower limit of the mapping interval to obtain the plurality of preprocessing coefficients and the plurality of first arrays.

3. The data fusion transmission method according to claim 2, characterized in that: The step of adjusting the plurality of intermediate arrays according to the mapping interval and the lower limit of the mapping interval to obtain the preprocessing coefficients and the plurality of first arrays includes: For each intermediate array, extract the maximum data and the minimum data as two preprocessing coefficients of the intermediate array, thereby obtaining a plurality of preprocessing coefficients; According to the first formula, the mapping interval and the lower limit of the mapping interval, the multiple intermediate arrays are adjusted to obtain the multiple first arrays, wherein the first formula: In the formula, is the mapping interval, is the lower limit of the mapping interval, The middle array data, is the maximum data of the middle array, is the minimum data of the middle array, The first array data, is the rounding function.

4. The data fusion transmission method according to claim 1, characterized in that: The first data fusion model pair includes a first encoding model and a first decoding model, and the using the multiple first arrays to adjust multiple parameters of the first data fusion model pair to obtain the second data fusion model pair includes: Obtaining a mapping interval, a first encoding model, and a first decoding model; Arrange the data appearing in the plurality of first arrays according to the magnitude of the frequencies of appearance in the plurality of first arrays to obtain a sample queue; Constructing a second data queue arranged in order of size and filling the mapping interval, wherein the data in the second data queue is an integer; constructing the second data queue as a label queue of the sample queue, wherein the label value of the sample queue data is negatively correlated with the occurrence frequency of the sample queue data in the plurality of first arrays; According to the sample queue and the label queue, multiple parameters of the first encoding model and multiple parameters of the first decoding model are adjusted, and the adjusted first encoding model and the adjusted first decoding model are used as a first data fusion model pair.

5. The data fusion transmission method according to claim 4, characterized in that: The adjusting the multiple parameters of the first encoding model and the multiple parameters of the first decoding model according to the sample queue and the label queue, and using the adjusted first encoding model and the adjusted first decoding model as a first data fusion model pair, includes: Randomly inputting a plurality of data of the sample queue into the first coding model; If the deviation between the output of the first coding model and the target label is greater than a first threshold, multiple parameters of the first coding model are adjusted according to the deviation between the output of the first coding model and the target label, and the process jumps to the step of randomly inputting multiple data of the sample queue into the first coding model, wherein the target label is a label corresponding to the data input into the first coding model; Randomly inputting a plurality of labels of the label queue into the first decoding model; If the deviation between the output of the first decoding model and the target data is greater than a second threshold, multiple parameters of the first decoding model are adjusted according to the deviation between the output of the first decoding model and the target data, and the process jumps to the step of randomly inputting multiple labels of the label queue into the first decoding model, wherein the target data is data of the sample queue identified by the label input into the first decoding model; The adjusted first encoding model and the adjusted first decoding model are used as the second encoding model and the second decoding model respectively.

6. The data fusion transmission method according to claim 5, characterized in that: The first coding model is: In the formula, is the output of the first encoding model, For the Line The output of the intermediate function is as follows, is the node meta-function, is the first bias parameter, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, For the The output of the input node, For the The weight parameters of the input nodes, For the Input data, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters; The first decoding model is: In the formula, is the output of the second model, For the Line The output of the intermediate function is as follows, is the second bias constant, The output function is weights, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameters, is the output of the input node, is the weight parameter of the input node, For the Line The output of the intermediate function is as follows, For the Line The middle function of the column weight parameter.

7. The data fusion transmission method according to claim 6, characterized in that: The node meta-function is: In the formula, is the input quantity, is a natural constant.

8. A data fusion transmission device, characterized in that: Used to implement the data fusion transmission method according to any one of claims 1 to 7, the data fusion transmission device comprises: A data stream acquisition module, used to acquire multiple data streams; A data adjustment module, used for cutting the multiple data streams, performing data overlap processing on the multiple arrays obtained by cutting, and obtaining preprocessing coefficients and multiple first arrays; A fusion model construction module, used to adjust multiple parameters of the first data fusion model pair using the multiple first arrays to obtain a second data fusion model pair, wherein the second data fusion model pair includes a second encoding model and a second decoding model; as well as, A data transmission module is used to input the multiple first arrays into the second encoding model to obtain multiple second arrays, and to transmit multiple target model parameters, the multiple preprocessing coefficients and the multiple second arrays, wherein the target model parameters are the parameters of the second decoding model.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in any one of claims 1 to 7 are implemented.

Images