Multi-source heterogeneous information fusion method based on WHO-RNN algorithm

Abstract

The invention proposes a multi-source heterogeneous information fusion method based on a WHO-RNN algorithm, and relates to the technical field of data fusion, and the method comprises the steps: S1, obtaining to-be-processed multi-source heterogeneous information, and carrying out the data preprocessing; s2, establishing an RNN network model based on a recurrent neural network; s3, optimizing network parameters of the RNN network model through a WHO algorithm to obtain a multi-source heterogeneous data fusion model; s4, inputting the preprocessed multi-source heterogeneous information into the improved multi-source heterogeneous data fusion model to complete multi-source heterogeneous information fusion; according to the method, the fusion effect of the multi-source heterogeneous information can be remarkably improved, the data integrity of the multi-source heterogeneous information is ensured, and compared with a traditional information management method, the method has remarkable data processing efficiency and guarantees the security of network data.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion, and in particular, to a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm. Background Art

[0002] With the continuous development of information fusion technology, the amount of communication information is increasing. It is necessary to perform fusion processing on communication information, which can effectively reduce the situation of data processing terminal paralysis caused by overloading of the data processing terminal due to excessive communication information. Although information fusion originated in the military field, the current research on it is not limited to this and has been extended to various fields, such as target recognition, geological science applications, medical applications, intelligent transportation, remote sensing systems, etc.

[0003] Although information fusion technology is widely used, there is still a lack of a systematic theoretical basis for information fusion technology. Most of the existing fusion algorithms for processing similar information cannot perfectly support the construction and fusion of different types of information models, and often have problems such as poor fault tolerance, incomplete data structure, and low detection accuracy. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm, which can significantly improve the fusion effect of multi-source heterogeneous information and ensure network data security.

[0005] The technical solution of the present invention is as follows:

[0006] In a first aspect, the present application provides a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm, which includes the following steps:

[0007] S1. Obtain multi-source heterogeneous information to be processed and perform data preprocessing;

[0008] S2. Establish an RNN network model based on a recurrent neural network;

[0009] S3. Optimize the network parameters of the RNN network model through the WHO algorithm to obtain a multi-source heterogeneous data fusion model;

[0010] S4. Input the preprocessed multi-source heterogeneous information into the improved multi-source heterogeneous data fusion model to complete multi-source heterogeneous information fusion.

[0011] Further, in step S1, the above preprocessing includes node positioning and mapping analysis;

[0012] Among them, the calculation process of the above node positioning includes:

[0013]

[0014] αX = β

[0015] X = (α T α) -1 α T β

[0016] where S Hop is, (x i , y i ), (x j , y j ) are the coordinates of node i and node j respectively, h j is the number of hop segments between node i and node j, α and β are both matrix parameters, X is node positioning, α is the transpose matrix;

[0017] The calculation process of the above mapping analysis includes:

[0018]

[0019] where σ is the mapping set, is the mapping from the quantization space in the fusion space, τ represents the whole process of the transformation of the original multi-source heterogeneous information into the data in the fusion space after the mapping, f is the mapping, is the spatial representation before the fusion of n multi-source heterogeneous data sets, is the state data existing in the space before the fusion of the m-th node of n multi-source heterogeneous data sets, m is the total number of nodes, θ t is the quantization space at time t, is the k-th heterogeneous data obtained by the data source l at the current time t during the fusion, is the data extreme value, τ(·) is the mapping relationship function, ρ is the finally obtained spatial representation, Δt is the time difference between the fusion data, δ is the number of data fusion times, is the data extreme value at time t.

[0020] Furthermore, in step S2, the calculation process of establishing the RNN network model includes:

[0021]

[0022] γ = 0.5×(ψ + O) + c

[0023] where O ξg is the g-th input data of the RNN network model, f(U ξ ) is the mapping of the g-th input data of the RNN network model, O ξ is the output data, ε is the error function, ξ is the mapping set, Υ is the number of neurons in the hidden layer, ψ is the number of neurons in the input layer, O is the number of neurons in the output layer, and c is a constant.

[0024] Further, in step S3, the calculation formula of the above WHO algorithm includes:

[0025]

[0026] In the formula, is the next position of the leader of the i-th group of populations, Z is the adaptation mechanism, R is a random number within the range of [-2, 2], W H is the water source position, and S Gi is the current position of the leader of the i-th group of populations.

[0027] Further, step S3 includes:

[0028] S31. Initialize the parameters of the RNN network model and the parameters of the WHO wild horse optimization algorithm;

[0029] S32. Input data samples into the RNN, optimize and solve the fitness values of wild horses in the population through the WHO algorithm, and determine the optimal value, the worst value and their corresponding wild horse positions;

[0030] S33. Sort all the wild horse fitness values in the optimization process, select the optimal wild horse to act as the leader of the new wild horse population, and modify the leader position again;

[0031] S34. Randomly select 10% of the wild horses from the wild horse population as foal wild horses, and modify the positions of the foal wild horses again;

[0032] S35. Update the optimal and worst positions and their corresponding fitness values of the existing wild horse population;

[0033] S36. Determine whether the iteration upper limit is reached. If so, output the optimal parameters to the RNN network model. If not, return to step S32;

[0034] S37. The RNN network model is trained based on the optimal parameters to obtain a multi-source heterogeneous data fusion model.

[0035] In a second aspect, the present application provides an electronic device, including:

[0036] A memory for storing one or more programs;

[0037] A processor;

[0038] When the above one or more programs are executed by the above processor, a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm as described in any one of the above first aspects is implemented.

[0039] In a third aspect, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm as described in any one of the above first aspects.

[0040] Compared with the prior art, the present invention has at least the following advantages or beneficial effects:

[0041] A multi-source heterogeneous information fusion method based on the WHO-RNN algorithm of the present invention optimizes the network parameters of the RNN network model by introducing the WHO algorithm, which can ensure global parameter optimization when the RNN performs multi-source heterogeneous data fusion, guarantee that the optimal parameters can be quickly found during model training, overcome the defects of low convergence efficiency and low accuracy faced by the RNN during the training process. At the same time, the diversity of the WHO algorithm itself ensures a large search margin for the RNN during the learning process, and can better avoid the RNN falling into local optimum compared with other algorithms, thus significantly improving the global search ability and the fusion effect of multi-source heterogeneous information, ensuring the data integrity of multi-source heterogeneous information, having significant data processing efficiency compared with traditional information management methods, and guaranteeing the security of network data. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0043] Figure 1 A multi-source heterogeneous information fusion method based on the WHO-RNN algorithm of the present invention;

[0044] Figure 2 A schematic structural block diagram of an electronic device according to an embodiment of the present invention.

[0045] Icons: 101, memory; 102, processor; 103, communication interface. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0047] It should be noted that, in this article, the term "comprise" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or equipment that includes a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment.

[0048] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0049] Example 1

[0050] See also Figure 1 , Figure 1 The figure shows a step diagram of a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm provided in an embodiment of the present application.

[0051] In a first aspect, the present application provides a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm, which comprises the following steps:

[0052] S1. Obtain multi-source heterogeneous information to be processed and perform data preprocessing;

[0053] S2. Establish an RNN network model based on recurrent neural network;

[0054] S3, optimizing the network parameters of the RNN network model through the WHO algorithm to obtain a multi-source heterogeneous data fusion model;

[0055] S4. Input the preprocessed multi-source heterogeneous information into the improved multi-source heterogeneous data fusion model to complete the multi-source heterogeneous information fusion.

[0056] As a preferred implementation, in step S1, preprocessing includes node positioning and mapping analysis;

[0057] The present invention uses the DV-Hop algorithm to perform node positioning calculation. First, the minimum sending data packet is set. After the latter is forwarded, the hop count of the node automatically increases by 1 and completes self-comparison after receiving the data packet, and then saves the data packet with the lowest hop value. After the lowest hop value is determined, all nodes obtain the corresponding lowest hop value. At this time, the average hop value can be expressed as:

[0058]

[0059] αX=β

[0060] Using the least squares method, the unknown node location can be solved:

[0061] X=(α T α)-1 α T β

[0062] In the formula, S Hop is, (x i , y i ), (x j , y j ) are the coordinates of node i and node j respectively, h j is the number of hops between node i and node j, α and β are both matrix parameters, X is node positioning, and α T is the transposed matrix;

[0063] The calculation process of mapping analysis includes:

[0064] Let the mapping set be:

[0065]

[0066] Then the spatial representation before the fusion of n multi-source heterogeneous data sets is:

[0067]

[0068] Among them, the rows of the matrix represent the spatial coverage targets before the fusion of multi-source heterogeneous data; the columns of the matrix represent their attributes; m represents the maximum characteristic quantity of the target, and if the target does not cover this characteristic, the value is set to 0; let the quantization space at the current t moment be:

[0069]

[0070] Obtain the representation formula of the mapping relationship function:

[0071]

[0072] If the data is preprocessed in advance during the data fusion process, the mapping relationship can be regarded as 1:1, and the finally obtained space can be represented as:

[0073]

[0074] In the formula, σ is the mapping set, is the mapping from the quantization space in the fusion space, τ represents the whole process of the transformation of the original multi-source heterogeneous information into the data in the fusion space after the mapping, f is the mapping, is the spatial representation before the fusion of n multi-source heterogeneous data sets, is the state data existing in the space before the fusion of the m-th node of n multi-source heterogeneous data sets, m is the total number of nodes, θ t is the quantization space at the t moment, is the k-th heterogeneous data obtained by the data source l at the current t moment during the fusion, Let it be the data extreme value, τ(·) be the mapping relation function, ρ be the finally obtained spatial representation, Δt be the time difference between the fused data, and δ be the number of data fusion times. It is the data extreme value at time t.

[0075] As a preferred implementation manner, in step S2, the calculation process of establishing the RNN network model includes:

[0076]

[0077] γ = 0.5×(ψ + O) + c

[0078] In the formula, O ξg is the g-th input data of the RNN network model, f(U ξ ) is the mapping of the g-th input data of the RNN network model, O ξ is the output data, ε is the error function, ξ is the mapping set, Υ is the number of neurons in the hidden layer, ψ is the number of neurons in the input layer, O is the number of neurons in the output layer, and c is a constant.

[0079] It should be noted that the RNN network model includes an input layer, a hidden layer, and an output layer; among them, the output layer determines the hidden layer after passing through the screening nodes, and the hidden layer expands and iterates to change the input value into the output value and outputs it to the output layer.

[0080] As a preferred implementation manner, in step S3, the calculation formula of the WHO algorithm includes:

[0081]

[0082] In the formula, is the next position of the leader of the i-th group of the population, Z is the adaptation mechanism, R is a random number within the range of [-2, 2], W H is the water source position, S Gi is the current position of the leader of the i-th group of the population.

[0083] As a preferred implementation manner, step S3 includes:

[0084] S31. Initialize the parameters of the RNN network model and the parameters of the WHO wild horse optimization algorithm;

[0085] S32. Input the data sample into the RNN, find the fitness value of the wild horses in the population by optimizing with the WHO algorithm, and determine the optimal value, the worst value, and their corresponding wild horse positions;

[0086] S33. Sort all the wild horse fitness values in the optimization process, select the optimal wild horse to act as the leader of the new wild horse population, and modify the leader position again;

[0087] S34. Randomly select 10% of the wild horses from the wild horse population as juvenile wild horses and modify the positions of the juvenile wild horses again;

[0088] S35. Update the optimal and worst positions and their corresponding fitness values of the existing wild horse population;

[0089] S36. Determine whether the iteration upper limit is reached. If so, output the optimal parameters to the RNN network model. If not, return to step S32;

[0090] S37. The RNN network model is trained based on the optimal parameters to obtain a multi-source heterogeneous data fusion model.

[0091] Embodiment 2

[0092] Please refer to Figure 2 , Figure 2 which is a schematic structural block diagram of an electronic device provided by an embodiment of the present application.

[0093] An electronic device includes a memory 101, a processor 102, and a communication interface 103. The memory 101, the processor 102, and the communication interface 103 are directly or indirectly electrically connected to each other to realize data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The memory 101 can be used to store software programs and modules. The processor 102 executes various functional applications and data processing by executing the software programs and modules stored in the memory 101. The communication interface 103 can be used to communicate with other node devices for signaling or data.

[0094] Among them, the memory 101 can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc.

[0095] The processor 102 can be an integrated circuit chip with signal processing capabilities. The processor 102 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0096] It can be understood that the structure shown in the figure is only schematic. A multi-source heterogeneous information fusion method based on the WHO-RNN algorithm may also include more or fewer components than those shown in the figure, or have a different configuration from that shown in the figure. Each component shown in the figure can be implemented using hardware, software, or a combination thereof.

[0097] In the embodiments provided in the present application, it should be understood that the disclosed method can also be implemented in other ways. The above-described embodiments are merely illustrative. For example, the flowcharts or block diagrams in the accompanying drawings show the possible architectures, functions, and operations of methods and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0098] In addition, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist alone, or two or more modules can be integrated to form an independent part.

[0099] When the above functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0100] The above are only the preferred embodiments of this application and are not intended to limit this application. For those skilled in the art, this application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included within the protection scope of this application.

[0101] For those skilled in the art, it is obvious that this application is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of this application, this application can be implemented in other specific forms. Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of this application is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within this application. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A multi-source heterogeneous information fusion method based on WHO-RNN algorithm, characterized in that: The following steps are involved: S1. Obtain multi-source heterogeneous information to be processed and perform data preprocessing; S2. Establish an RNN network model based on recurrent neural network; S3, optimizing the network parameters of the RNN network model through the WHO algorithm to obtain a multi-source heterogeneous data fusion model; S4. Input the preprocessed multi-source heterogeneous information into the improved multi-source heterogeneous data fusion model to complete the multi-source heterogeneous information fusion.

2. A multi-source heterogeneous information fusion method based on WHO-RNN algorithm as claimed in claim 1, characterized in that: In step S1, the preprocessing includes node positioning and mapping analysis; The node location calculation process includes: αX＝β X＝(α T α) -1 α T β In the formula, S Hop For, (x i ,y i )、(x j ,y j ) are the coordinates of node i and node j, h j is the number of hops between node i and node j, α and β are matrix parameters, X is the node location, α T is the transposed matrix; The calculation process of the mapping analysis includes: In the formula, σ is the mapping set, is the mapping from the quantized space in the fusion space, τ represents the entire process of transforming the original multi-source heterogeneous information into the data in the fusion space after the mapping is completed, f is the mapping, It is the spatial representation before the fusion of n multi-source heterogeneous data sets. is the state data of the space before the mth node fusion of n multi-source heterogeneous data sets, m is the total number of nodes, θ t is the quantization space at time t, θs lk is the kth heterogeneous data obtained by fusion of data source l at the current time t, is the data extreme value, τ(·) is the mapping relationship function, ρ is the final spatial representation, Δt is the time difference between the fused data, δ is the number of data fusion times, is the data extreme value at time t.

3. The multi-source heterogeneous information fusion method based on WHO-RNN algorithm as claimed in claim 1, characterized in that: In step S2, the calculation process of establishing the RNN network model includes: γ=0.5×(ψ+O)+c In the formula, O ξg is the g-th input data of the RNN network model, f(U ξ ) is the mapping of the g-th input data of the RNN network model, O ξ is the output data, ε is the error function, ξ is the mapping set, γ is the number of neurons in the hidden layer, ψ is the number of neurons in the input layer, O is the number of neurons in the output layer, and c is a constant.

4. The multi-source heterogeneous information fusion method based on WHO-RNN algorithm as claimed in claim 1, characterized in that: In step S3, the calculation formula of the WHO algorithm includes: In the formula, is the next position of the leader of the i-th group, Z is the adaptive mechanism, R is a random number in the range of [-2,2], and W H is the water source location, is the current position of the leader of the i-th group.

5. The multi-source heterogeneous information fusion method based on WHO-RNN algorithm as claimed in claim 1, characterized in that: Step S3 includes: S31, initializing the parameters of the RNN network model and the parameters of the WHO wild horse optimization algorithm; S32, input data samples to RNN, use the WHO algorithm to find the fitness value of wild horses in the population, and determine the optimal value, the worst value and their corresponding wild horse points; S33, sorting the fitness values ​​of all wild horses in the optimization process, selecting the best wild horse to serve as the new leader of the wild horse population, and re-modifying the leader position; S34, randomly selecting 10% of the wild horses from the wild horse population as young wild horses, and re-modifying the positions of the young wild horses; S35, updating the optimal and the worst positions of the existing wild horse population and their corresponding fitness values; S36, determining whether the iteration limit is reached, if so, outputting the optimal parameters to the RNN network model, if not, returning to step S32; S37. The RNN network model is trained based on the optimal parameters to obtain a multi-source heterogeneous data fusion model.

6. An electronic device, characterized in that: include: A memory for storing one or more programs; processor; When the one or more programs are executed by the processor, a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm as described in any one of claims 1 to 5 is implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, a multi-source heterogeneous information fusion method based on the WHO-RNN algorithm as described in any one of claims 1 to 5 is implemented.

Images