Fusion method and system based on wide and narrow band data of multi-source sensors and storage medium

Abstract

The application belongs to the technical field of sensor data analysis, and particularly relates to a fusion method and system based on wide and narrow band data of multiple source sensors and a storage medium, which comprises the following steps: in a fusion period T, the measurement value of a sensor is collected and initialized, the parameters of an algorithm are determined, the current position vector and velocity vector of a particle are randomly initialized in a search range, the target state at the moment is estimated and fused in turn to obtain the local estimation of each sensor; the local estimation of each sensor is transmitted to a fusion center, the data of the sensor node is processed by the fusion center; and the estimation value of the target state based on the global estimation information is obtained when the fusion period T arrives. The application enables the information fusion to be more accurately executed, and enables the fusion to be performed in an approximate optimal manner on the sensor data under the conditions that the reliability of the information source, the redundancy and complementarity of the information and the hierarchical structure uncertainty of the fusion are uncertain.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sensor data analysis, and particularly relates to a fusion method and system based on wideband and narrowband data of multi-source sensors and a storage medium. BACKGROUND

[0002] With the development of modern technology, more and more attention is paid to obtaining more comprehensive, accurate and reliable environmental situation information by using multi-sensor data fusion technology. Among them, the distributed estimation data fusion method has developed rapidly and is widely used in many fields such as regional monitoring and target positioning because of its low resource consumption, strong scalability and good robustness. However, most of the existing distributed estimation data fusion methods are designed for synchronous sensor networks, that is, the sampling periods of multiple sensors are required to be the same, the initial deviations are required to be the same, and there is no communication delay. In fact, these requirements are difficult to meet in practice.

[0003] At present, many power distribution stations are equipped with equipment operation state monitoring devices, but the monitoring system in the station house often only has data display function, lacks data analysis algorithm, and cannot intuitively display the actual situation of the station house. In addition, the wideband data and narrowband sensor data in the station house are difficult to fuse, the gateway communication bandwidth is insufficient, the local computing capacity is poor, a large amount of data cannot be analyzed locally, and the utilization rate of the gateway northbound channel resources is greatly reduced.

[0004] The information disclosed in this BACKGROUND section is only for the purpose of enhancing the understanding of the background of the present application and should not be taken as admitting that such information is prior art with respect to the present application. SUMMARY

[0005] The purpose of the present application is to provide a fusion method and system based on wideband and narrowband data of multi-source sensors and a storage medium to solve the problems in the background art.

[0006] The fusion method and system based on wideband and narrowband data of multi-source sensors and the storage medium provided by the present application are applied to the equipment operation state monitoring system in the power distribution station. The algorithm solves the problem of how to fuse the wideband and narrowband data of the sensors to intuitively display the actual situation of the station house when the monitoring system in the station house only has data display function and lacks data analysis algorithm.

[0007] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0008] The fusion method based on wideband and narrowband data of multi-source sensors comprises the following steps:

[0009] S1: in the fusion cycle T, collect the measurement of the sensor and initialize it, determine the parameters of the algorithm, randomly initialize the current position vector and velocity vector of the particle in the search range, where the dimension d of the particle is the number N of the sensor to be selected, estimate and fuse the target state at the moment in turn to obtain the local estimation of each sensor;

[0010] S2: transmit the local estimation of each sensor to the fusion center, the fusion center processes the data of the sensor node, fuses the data from the sensor node into a current time point, and calculates the probability of using each sensor, and calculates the fitness value of each particle according to the objective function;

[0011] S3: the fusion center returns the position at the current time to each sensor node, and the sensor node takes it as the initial value at the next time;

[0012] S4: when the fusion cycle T reaches the target state, the estimation value based on the global estimation information is obtained.

[0013] Further, the processing step in step S2 is:

[0014] 1) compare the fitness value of each particle with the best value it has experienced, if it is better, take it as the individual optimal value of the current particle;

[0015] 2) compare the individual optimal value of each particle with the group optimal value, if it is better, take it as the group optimal value;

[0016] 3) update the speed and position of each particle, if the termination condition is not reached, go to the calculation of the probability of using each sensor again, and calculate the fitness value of each particle according to the objective function.

[0017] Further, the mathematical formula of the particle swarm algorithm can be expressed as:

[0018] θ jd = ωθ jd + c1rand1()(p jd -x jd ) + c2rand2()(p kd -x jd ) (1),

[0019] Where x jd represents the current position vector of the particle, θ jd represents the motion vector of the particle, p jd represents the individual position optimal value of the particle, and p kddenoted as the optimal value of the population, j = 1, 2, ..., m indicates that the population consists of m particles, k represents the number of optimal values ​​of individual positions in the population, d = 1, 2, ..., D represents the d-dimensional space, ω represents the inertial weight, c1 and c2 represent non-negative learning constants, and rand1() and rand2() represent random numbers between [0, 1].

[0020] Furthermore, to achieve convergence of equation (1), the specific data process includes: updating the particle state:

[0021] L=Z|x jd (t)-p t |,

[0022] x jd (t+1)=p d -Lln(1 / u),

[0023] Where p t p represents the center of potential. d The mean is the optimal point, and t represents less than. . a nonnegative constant.

[0024] Furthermore, in step S1, multiple sensors are used to track the target. Each sensor has a different communication delay, and the measurement equation for each sensor can be expressed as:

[0025]

[0026] Let represent the measurement vector of the i-th sensor at each time step. Represents the measurement matrix. Indicates measurement noise, x k This represents the state vector at time k.

[0027] Furthermore, when the fusion center performs processing, it first obtains the state x at time K-1. k-1 Based on global estimate and the corresponding error covariance P k-1,k-1 The specific formula is as follows:

[0028]

[0029]

[0030] Q represents the system transition matrix from time k-1 to time k; k-1 Let be the covariance matrix.

[0031] Furthermore, as time progresses forward, the process is recursively applied to each of the following: The solution is performed to obtain t. iTime state estimation And corresponding error covariance Through iteration, the k time is obtained, and then the global estimation and estimation error covariance are obtained by using the fusion algorithm, and the specific formula is:

[0032]

[0033]

[0034] Further, the fusion center obtains the measurement information of each sensor node Substituted into the solution The state estimation And corresponding error covariance Get

[0035] The second object of the application is to provide a fusion system based on multi-source sensor wide and narrow band data, which is used to realize the fusion method based on multi-source sensor wide and narrow band data.

[0036] A fusion system based on multi-source sensor wide and narrow band data, comprising:

[0037] Memory for storing computer programs;

[0038] The processor is used to execute the computer program to realize the steps of the fusion method based on multi-source sensor wide and narrow band data.

[0039] The third object of the application is to provide a storage medium for carrying the computer program of the fusion method based on multi-source sensor wide and narrow band data, which has the same technical effect.

[0040] A storage medium, the storage medium has a computer program stored thereon, the computer program can be read out by a computer, and the computer program is executed to realize the steps of the fusion method based on multi-source sensor wide and narrow band data.

[0041] Compared with the prior art, the application has the following beneficial effects:

[0042] The application obtains local estimation of each sensor by fusing the measured values observed by each sensor within a fusion cycle T, then fuses the data from the sensor nodes into a current time point at the fusion center, calculates the estimation value of the target state based on the global estimation information at the time when the fusion cycle T arrives, and fuses the broadband data such as video and image and the narrowband sensor data in the current station, so that the actual situation in the station can be intuitively displayed. BRIEF DESCRIPTION OF DRAWINGS

[0043] Fig. 1 Method flow chart of the fusion method based on wide and narrow band data of multi-source sensors in the application;

[0044] Fig. 2 Data fusion structure of the fusion method based on wide and narrow band data of multi-source sensors in the application;

[0045] Fig. 3 Method flow chart of the fusion method based on wide and narrow band data of multi-source sensors in the application. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application.

[0047] Referring to Figs. 1-3 The fusion method based on wide and narrow band data of multi-source sensors includes the following steps:

[0048] S1: within a fusion cycle T, the measured values of the sensors are collected and initialized, the parameters of the algorithm are determined, the current position vector and velocity vector of the particles are randomly initialized in the search range, wherein the dimension d of the particles is the number N of the sensors to be selected, the target state at this time is estimated and fused in turn to obtain the local estimation of each sensor;

[0049] S2: the local estimation of each sensor is transmitted to the fusion center, the fusion center processes the data of the sensor nodes, fuses the data from the sensor nodes into a current time point, and calculates the probability of use of each sensor, and calculates the fitness value of each particle according to the target function formula;

[0050] S3: the fusion center returns the current time position to each sensor node, and the sensor node takes it as the initial value at the next time;

[0051] S4: obtaining an estimated value of the target state based on global estimation information when the fusion cycle T moment arrives.

[0052] The application obtains the local estimation of each sensor by fusing the measured values observed by each sensor within the fusion cycle T, and then fuses the data from the sensor nodes into one current time point in the fusion center, and obtains the estimated value of the target state based on the global estimation information when the fusion cycle T moment arrives, so as to fuse the broadband data such as video and image and the narrowband sensor data in the station building, and make the actual situation in the station building be intuitively displayed.

[0053] In step S2, the particle swarm algorithm is adopted, the search process in the optimization space is realized through mutual cooperation and competition between different particles in the group to find the group optimal value, and then the fitness value of each particle is calculated. The specific processing steps are as follows:

[0054] 1) comparing the fitness value of each particle with the best value experienced by it, if better, taking it as the individual optimal value of the current particle;

[0055] 2) comparing the individual optimal value of each particle with the group optimal value, if better, taking it as the group optimal value;

[0056] 3) updating the speed and position of each particle, if the termination condition is not reached, returning to the calculation of the probability of use of each sensor, and calculating the fitness value of each particle according to the target function formula. It should be noted that the termination condition is to reach the maximum iteration number or the group optimal value meets the minimum limit.

[0057] Specifically, the mathematical formula of the particle swarm algorithm can be expressed as:

[0058] θ jd = ωθ jd + c1rand1()(p jd -x jd ) + c2rand2()(p kd -x jd ) (1),

[0059] Wherein x jd represents the current position vector of the particle, θ jd represents the motion vector of the particle, p jd represents the individual position optimal value of the particle, and p kdrepresents the global optimal value of the group, j = 1, 2, …, m represents that the colony is composed of m particles, k represents the number of individual position optimal values in the group, d = 1, 2, …, D represents a d-dimensional space, ω represents an inertia weight, which is a non-negative number, and when the value of ω is large, the global optimization ability is strong and the local optimization ability is weak. Conversely, the local optimization ability is strong. c1 and c2 represent non-negative learning constants, if c1 = 0, it is easy to fall into a local optimal solution and cannot jump out; if c2 = 0, it is easy to cause the algorithm to converge slowly; and when c1 and c2 are not 0, it is easier to keep the balance of convergence speed and search effect, which is a better choice. rand1() and rand2() represent random numbers between 0 and 1.

[0060] More specifically, the convergence of formula (1) is carried out, and the specific data process includes: carrying out particle state updating:

[0061] L = Z | x jd (t) - p t |,

[0062] x jd (t + 1) = pd - Lln(1 / u),

[0063] where p t represents a potential center point, p d represents a mean optimal point, and t represents a non-negative constant less than .

[0064] Please continue to refer to step S1, in which a plurality of sensors are used to track the target, each sensor has different communication delay, and the measurement equation of each sensor can be expressed as:

[0065]

[0066] xi(t) represents the measurement vector of the i-th sensor at each time, represents a measurement matrix, represents measurement noise, and x k represents a state vector at time k.

[0067] Specifically, when the fusion center processes, first, the state x k-1 at time K-1 is obtained based on the global estimation value k-1,k-1 and the corresponding error covariance P .

[0068]

[0069]

[0070] This represents the system transition matrix from time k-1 to time k; O k-1 Let be the covariance matrix.

[0071] Furthermore, as time progresses forward, the process is recursively applied to each of the following: The solution is performed to obtain t. i Time-state estimation and the corresponding error covariance The time step k is obtained through iteration, and then the global estimate and the estimation error covariance are obtained using a fusion algorithm. The specific formula is as follows:

[0072]

[0073]

[0074] Based on this, the fusion center will obtain the measurement information from each sensor node. Substitute into the solution In this process, the state estimate at the corresponding time is obtained. and the corresponding error covariance get

[0075] The present invention also provides a fusion system based on wideband and narrowband data from multiple sensors, comprising:

[0076] Memory, used to store computer programs;

[0077] A processor is used to implement the steps of the above-described method for fusing wide and narrow band data from multiple sensors when executing the computer program.

[0078] The present invention also provides a storage medium for carrying a computer program for a fusion method based on multi-source sensor broadband and narrow-band data, achieving the same technical effect. The storage medium stores the computer program, which can be read by a computer. When executed, the computer program implements the steps of the aforementioned fusion method based on multi-source sensor broadband and narrow-band data.

[0079] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0080] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Fig. 1 one or more flowcharts and / or blocks Fig. 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0081] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Fig. 1 one or more flowcharts and / or blocks Fig. 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0082] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Fig. 1 one or more flowcharts and / or blocks Fig. 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0083] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments described and shown, and it is therefore intended that the application be limited only by the scope of the claims presented with this written description.

[0084] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the claims, the application can be practiced otherwise than as specifically described herein.

[0085] The above merely describes preferred embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can make equivalent replacements or changes within the technical scope disclosed by the present application and according to the technical solutions and inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A fusion method based on multi-source sensor wide and narrow band data, characterized in that, The method comprises the following steps: S1: in a fusion cycle T, collecting and initializing the measurement values of the sensors, determining the parameters of the algorithm, randomly initializing the current position vector and velocity vector of the particles in the search range, wherein the dimension d of the particles is the number N of the sensors to be selected, and sequentially estimating and fusing the target state to obtain the local estimation of each sensor; S2: transmitting the local estimation of each sensor to the fusion center, the fusion center processing the data of the sensor nodes, fusing the data from the sensor nodes into a current time point, and calculating the probability of using each sensor, and calculating the fitness value of each particle according to the target function formula; S3: the fusion center returns the position of the current time to each sensor node, and the sensor node takes it as the initial value of the next time; S4: when the fusion cycle T arrives, the estimation value of the target state based on the global estimation information is obtained; The fusion center, in processing, first obtains the state at time K-1 based on the global estimate and the corresponding error covariance The specific formula is: ， ， denotes the system transition matrix from time k-1 to time k; is the covariance matrix; As time goes on, the state estimation of k is obtained by recursively solving and the corresponding error covariance , and the global estimation and estimation error covariance are obtained by iteration at time k, and the specific formula is as follows:​​ ； The fusion center will get the measurement information of each sensor node Substitute into the solution Get the state estimation at the corresponding time And the corresponding error covariance Get .

2. The fusion method based on multi-source sensor wide and narrow band data according to claim 1, characterized in that, The processing steps in step S2 are: 1) compare the fitness value of each particle with the best value it has experienced, if it is better, take it as the individual optimal value of the current particle; 2) compare the individual optimal value of each particle with the group optimal value, if it is better, take it as the group optimal value; 3) update the speed and position of each particle, and if the termination condition is not reached, go back to calculating the probability of using each sensor, and calculating the fitness value of each particle according to the target function formula.

3. The fusion method based on multi-source sensor wide and narrow band data according to claim 2, characterized in that, The mathematical formula of the particle swarm algorithm can be expressed as: （1）， wherein represents the current position vector of the particle, represents the motion vector of the particle, represents the individual position optimum of the particle, represents the swarm optimum, represents that the swarm consists of m particles, k represents the number of individual position optimum in the swarm, represents the d-dimensional space, represents the inertia weight, represents the non-negative learning constant, represents a random number between and 4. The fusion method based on multi-source sensor wide and narrow band data according to claim 3, characterized in that, Converge to formula (1), and the specific data process includes: updating the particle state: wherein represents a potential center point, represents a mean optimum point, t represents a non-negative constant smaller than 1.

5. The method of claim 1, wherein the method is based on a fusion of multi-source sensor wide and narrow band data. In step S1, multiple sensors are used to track the target, each sensor has different communication delay, and the measurement equation of each sensor can be expressed as: ， xi represents the measurement vector of the i-th sensor at each time instant, xi represents the measurement matrix, xi represents the measurement noise, xi represents the state vector at time instant k.

6. A fusion system based on multi-source sensor wide and narrow band data, characterized in that, It comprises: A memory for storing a computer program; A processor for executing the computer program to realize the steps of the fusion method based on wide and narrow band data of multiple source sensors according to any one of claims 1-5.

7. A storage medium, characterized by The storage medium stores a computer program, which can be read out by a computer, and the computer program is executed to realize the steps of the fusion method based on wide and narrow band data of multiple source sensors according to any one of claims 1-5.

Images