CN117932533A - Emotion science multi-source data fusion method and system based on Bayesian statistics - Google Patents
Emotion science multi-source data fusion method and system based on Bayesian statistics Download PDFInfo
- Publication number
- CN117932533A CN117932533A CN202410064329.8A CN202410064329A CN117932533A CN 117932533 A CN117932533 A CN 117932533A CN 202410064329 A CN202410064329 A CN 202410064329A CN 117932533 A CN117932533 A CN 117932533A
- Authority
- CN
- China
- Prior art keywords
- source data
- data
- decision
- information
- prior
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013477 bayesian statistics method Methods 0.000 title claims abstract description 24
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 16
- 230000008451 emotion Effects 0.000 title description 2
- 238000009826 distribution Methods 0.000 claims abstract description 104
- 230000004927 fusion Effects 0.000 claims abstract description 61
- 238000011160 research Methods 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims description 20
- 238000000034 method Methods 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 12
- 239000006185 dispersion Substances 0.000 claims description 9
- 238000013507 mapping Methods 0.000 claims description 9
- 230000007704 transition Effects 0.000 claims description 6
- 241000772415 Neovison vison Species 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012067 mathematical method Methods 0.000 claims description 2
- 238000005553 drilling Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of multi-source data fusion, in particular to an earth science multi-source data fusion method and system based on Bayesian statistics, comprising the following steps: constructing a probability model which fuses the multi-source data and is used for quantitatively deciding a prediction problem and an uncertainty problem in the geoscience research through a Bayesian statistical framework to obtain a multi-source data fusion decision model; and inputting the prior information of the mathematical multisource data and the observation information of the multisource data into a multisource data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision-making geoscience research. The invention can effectively fuse multi-source data, convert some priori information which is difficult to mathematical into various probability distributions, effectively improve the data utilization efficiency and improve the decision accuracy of the prediction problem and the uncertainty problem in the earth science research.
Description
Technical Field
The invention relates to the technical field of multi-source data fusion, in particular to an earth science multi-source data fusion method and system based on Bayesian statistics.
Background
Geoscience is a typical multi-source data discipline, and therefore sources of a priori information are also diverse. Taking geophysical data inversion as an example, besides defining the distribution type of the observed data, the constraint can be introduced by the experience of geologist, geological horizon information, drilling logging data, different geophysical parameter data and the like. These data are not only of diverse origin, but also vary widely in data type. For example, lithology data is usually classified data, drilling and logging data is usually continuous data, the experience of geologist is more difficult to mathematical, and how to statistically integrate the information determines the solving effect of prediction problems and uncertainty problems in the earth science.
In the prior art, data information is usually processed respectively and then manually interpreted, and the method is time-consuming and labor-consuming and cannot obtain better results.
Disclosure of Invention
The invention aims to provide a Bayesian statistics-based geoscience multi-source data fusion method to solve the technical problems that in the prior art, manual interpretation is time-consuming and labor-consuming, and a good result cannot be obtained.
In order to solve the technical problems, the invention specifically provides the following technical scheme:
a Bayesian statistics-based geoscience multi-source data fusion method comprises the following steps:
Determining multi-source data in a ground-based personalized teleoperation of the geoscience study, wherein the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
Acquiring prior information of multi-source data and observation information of the multi-source data;
constructing a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining prior information representing the multi-source data and a multi-source data fusion decision model of mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
and inputting the prior information of the mathematical multisource data and the observation information of the multisource data into a multisource data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision-making geoscience research.
As a preferable scheme of the invention, the construction method of the multi-source data fusion decision model comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
As a preferable scheme of the invention, the decision probability is obtained by carrying out weighted average on the posterior distribution through the decision weight of prior information of the multi-source data and the decision weight of observation information of the multi-source data.
As a preferred embodiment of the present invention, the method for determining the decision weight of the prior information of the multi-source data includes:
Acquiring prior information of multi-source data and observation information of the multi-source data;
Calculating the dispersion of prior information of the multi-source data and observation information of the multi-source data, and quantifying the instability of various data in the multi-source data by using the dispersion, wherein the dispersion is calculated by using a similarity function, and the instability is as follows:
Ki=Dis(mold,i,mnew,i);
Wherein K i is instability of ith data in the multi-source data, m old,i is priori information of the ith data in the multi-source data, m new,i is observation information of the ith data in the multi-source data, and Dis (m old,i,mnew,i) is Euclidean distance between m old,i and m new,i;
Carrying out min-max standardization processing on the instability, and linearly mapping the instability to the range of the [0,1] interval, wherein K i after the min-max standardization processing is as follows: k i'=(Ki -minK)/(maxK-minK), wherein minK is the minimum value of the instability of n data in the multi-source data, and max is the maximum value of the instability of n data in the multi-source data;
according to the instability of the multi-source data, constructing decision weights of prior information of the multi-source data which are adaptively adjusted according to the instability, wherein the decision weights of the prior information of the multi-source data are as follows:
Wi=-(1-Ki')^r*log Ki’;
Wherein W i is the decision weight of priori information of ith data in the multi-source data, K i' is the instability of the ith data in the multi-source data after min-max standardization processing, i is a counting variable, r is greater than 1, and r is the constant power.
As a preferred embodiment of the present invention, the method for determining decision weights of observation information of multi-source data includes:
Obtaining the decision weight of the observation information of the multi-source data according to the decision weight of the prior information of the multi-source data;
the decision weight of the observation information of the multi-source data is as follows:
Mi=1-Wi;
Wherein M i is the decision weight of the observation information of the ith data in the multi-source data, W i is the decision weight of the prior information of the ith data in the multi-source data, and i is the counting variable.
As a preferred embodiment of the present invention, the method for mathematically relating a priori information of multi-source data includes:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
As a preferred scheme of the invention, the invention provides a Bayesian statistics-based geoscience multi-source data fusion system, which is applied to the Bayesian statistics-based geoscience multi-source data fusion method, and comprises the following steps:
The data preprocessing module is used for determining multi-source data in the ground physical teleoperation of the earth science research, and the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
The data acquisition module is used for acquiring prior information of the multi-source data and observation information of the multi-source data;
The model building module is used for building a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining a multi-source data fusion decision model which characterizes prior information of the multi-source data and mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
The mathematical processing module is used for carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
The decision output module is used for inputting the prior information of the mathematical multi-source data and the observation information of the multi-source data into the multi-source data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision geoscience research.
As a preferred scheme of the invention, the model building module builds a multi-source data fusion decision model, which comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
As a preferable scheme of the invention, the model building module obtains the decision probability by carrying out weighted average on the posterior distribution through the decision weight of prior information of the multi-source data and the decision weight of observation information of the multi-source data.
As a preferred solution of the present invention, the mathematical processing module performs mathematical processing on a priori information of the multi-source data, including:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
Compared with the prior art, the invention has the following beneficial effects:
According to the invention, a Bayesian statistical method is introduced in the aspect of fusion of the geoscience multisource data, the multisource data can be effectively fused, a plurality of priori information which is difficult to mathematical is converted into various probability distributions, the data utilization efficiency is effectively improved, and the decision accuracy of the prediction problem and the uncertainty problem in the geoscience research is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It will be apparent to those of ordinary skill in the art that the drawings in the following description are exemplary only and that other implementations can be obtained from the extensions of the drawings provided without inventive effort.
FIG. 1 is a flowchart of a method for geoscience multisource data fusion provided by an embodiment of the present invention;
Fig. 2 is a block diagram of an geoscience multi-source data fusion system provided by an embodiment of the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the invention provides a bayesian statistics-based geoscience multi-source data fusion method, which comprises the following steps:
Determining multi-source data in a ground-based personalized teleoperation of the geoscience study, wherein the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
Acquiring prior information of multi-source data and observation information of the multi-source data;
constructing a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining prior information representing the multi-source data and a multi-source data fusion decision model of mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
and inputting the prior information of the mathematical multisource data and the observation information of the multisource data into a multisource data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision-making geoscience research.
In order to realize quantitative decision of the prediction problem and the uncertainty problem in the geoscience research based on the multi-source data, a probability model is constructed by utilizing Bayesian statistics, data of different sources (namely multi-source data from the field of earth materialization tele) are fused, posterior probability distribution calculation and analysis are performed by utilizing the probability model formed by Bayesian formulas (namely the multi-source data fusion decision model), and decision probability is obtained, so that the prediction problem and the uncertainty problem in the geoscience are solved.
The multi-source data fusion decision model obtains decision probability according to prior information and observation information of multi-source data, mathematics experience (prior information) of geological personnel, combines current actual conditions (observation information), improves data utilization efficiency, promotes fusion of the multi-source data, improves stability of Bayesian classification, and enhances decision accuracy.
Furthermore, in the multi-source data fusion decision model established based on the Bayesian formula, the decision force or speaking right (the capability of influencing posterior distribution) of prior information and observation information of the multi-source data is adaptively quantized, so that the accuracy of the multi-source data fusion decision model is maintained, and the robustness of the multi-source data fusion decision model is enhanced.
Specifically, the decision force of the prior information of the multi-source data is related to the time sequence discreteness of the multi-source data, the time sequence discreteness of the multi-source data reflects the stability degree of the multi-source data, the worse the stability degree (the higher the time sequence discreteness is, the higher the instability value is), the worse the reliability of the prior information is indicated, and therefore the smaller the influence capacity on posterior distribution is supposed to be, namely the smaller the decision force is, and on the contrary, the better the stability degree (the lower the time sequence discreteness is, the lower the instability value is), the higher the reliability of the prior information is indicated, and the larger the influence capacity on posterior distribution is supposed to be, namely the larger the decision force is supposed to be.
In summary, the invention constructs the self-adaptive adjustment of the prior information of the multi-source data according to the stability of the data to endow the prior information of the multi-source data with decision force, so that the higher the instability is, the smaller the decision weight is, the smaller the capability of influencing decision results is, and the accuracy of the decision probability in the multi-source data fusion decision model is ensured.
Similarly, the more unstable the multi-source data is, the lower the dependence on priori information is, and the more the current observation information is relied on, the optimal decision accuracy can be obtained.
According to the invention, the decision weight is adaptively adjusted in the multi-source data fusion, so that the quantitative decision of the multi-source data in the field of materialization remote of the fusion site is realized, and meanwhile, the accuracy of the quantitative decision of the multi-source data is ensured.
Multisource data includes logging information, lithology information, geological knowledge, and the like, wherein lithology data is typically of the categorical data type, drilling logging data is typically of the continuous data type, and geological knowledge is typically of the empirical data type.
In order to realize quantitative decision on the prediction problem and the uncertainty problem in the geoscience research based on multi-source data, a probability model is constructed by utilizing Bayesian statistics, data of different sources (namely multi-source data from the field of earth materialization tele) are fused, posterior probability distribution calculation and analysis are carried out by utilizing the probability model formed by Bayesian formulas (namely a multi-source data fusion decision model), and decision probability is obtained, so that the prediction problem and the uncertainty problem in the geoscience are solved, and the method specifically comprises the following steps:
The construction method of the multi-source data fusion decision model comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
In essence, d is observation information, and a relation can be established between d and m through a linear or nonlinear relation, and an operator of the relation is assumed to be G, and in general, observation data dζ Gm and p (m|d) express posterior distribution of decision probability of prediction problem and uncertainty problem in the geoscience research under the condition of acquiring the observation data d, wherein the distribution of m is calculated.
And carrying out weighted average on the posterior distribution through the decision weight of the prior information of the multi-source data and the decision weight of the observation information of the multi-source data to obtain the decision probability.
In the multi-source data fusion decision model established based on a Bayesian formula, the decision force or speaking right (the capability of influencing posterior distribution) of prior information and observation information of the multi-source data is adaptively quantized, so that the accuracy of the multi-source data fusion decision model is maintained, and the robustness of the multi-source data fusion decision model is enhanced, and the method specifically comprises the following steps:
the method for determining the decision weight of the prior information of the multi-source data comprises the following steps:
Acquiring prior information of multi-source data and observation information of the multi-source data;
Calculating the dispersion of prior information of the multi-source data and observation information of the multi-source data, and quantifying the instability of various data in the multi-source data by using the dispersion, wherein the dispersion is calculated by using a similarity function, and the instability is as follows:
Ki=Dis(mold,i,mnew,i);
Wherein K i is instability of ith data in the multi-source data, m old,i is priori information of the ith data in the multi-source data, m new,i is observation information of the ith data in the multi-source data, and Dis (m old,i,mnew,i) is Euclidean distance between m old,i and m new,i;
Carrying out min-max standardization processing on the instability, and linearly mapping the instability to the range of the [0,1] interval, wherein K i after the min-max standardization processing is as follows: k i'=(Ki -minK)/(maxK-minK), wherein minK is the minimum value of the instability of n data in the multi-source data, and max is the maximum value of the instability of n data in the multi-source data;
according to the instability of the multi-source data, constructing decision weights of prior information of the multi-source data which are adaptively adjusted according to the instability, wherein the decision weights of the prior information of the multi-source data are as follows:
Wi=-(1-Ki')^r*log Ki’;
Wherein W i is the decision weight of priori information of ith data in the multi-source data, K i' is the instability of the ith data in the multi-source data after min-max standardization processing, i is a counting variable, r is greater than 1, and r is the constant power.
The method for determining the decision weight of the observation information of the multi-source data comprises the following steps:
Obtaining the decision weight of the observation information of the multi-source data according to the decision weight of the prior information of the multi-source data;
the decision weight of the observation information of the multi-source data is as follows:
Mi=1-Wi;
Wherein M i is the decision weight of the observation information of the ith data in the multi-source data, W i is the decision weight of the prior information of the ith data in the multi-source data, and i is the counting variable.
Specifically, the decision force of the prior information of the multi-source data is related to the time sequence discreteness of the multi-source data, the time sequence discreteness of the multi-source data reflects the stability degree of the multi-source data, the worse the stability degree (the higher the time sequence discreteness is, the higher the instability value is), the worse the reliability of the prior information is indicated, and therefore the smaller the influence capacity on posterior distribution is supposed to be, namely the smaller the decision force is, and on the contrary, the better the stability degree (the lower the time sequence discreteness is, the lower the instability value is), the higher the reliability of the prior information is indicated, and the larger the influence capacity on posterior distribution is supposed to be, namely the larger the decision force is supposed to be.
In summary, the invention constructs the self-adaptive adjustment of the prior information of the multi-source data according to the stability of the data to endow the prior information of the multi-source data with decision force, so that the higher the instability is, the smaller the decision weight is, the smaller the capability of influencing decision results is, and the accuracy of the decision probability in the multi-source data fusion decision model is ensured.
Similarly, the more unstable the multi-source data is, the lower the dependence on priori information is, and the more the current observation information is relied on, the optimal decision accuracy can be obtained.
In order to realize the fusion effect of the multi-source data, the invention carries out mathematical treatment on the multi-source data, and specifically comprises the following steps:
the mathematical method for the prior information of the multi-source data comprises the following steps:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
As shown in fig. 2, the present invention provides a bayesian statistics-based geoscience multi-source data fusion system, which is applied to the bayesian statistics-based geoscience multi-source data fusion method, and the geoscience multi-source data fusion system includes:
The data preprocessing module is used for determining multi-source data in the ground physical teleoperation of the earth science research, and the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
The data acquisition module is used for acquiring prior information of the multi-source data and observation information of the multi-source data;
The model building module is used for building a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining a multi-source data fusion decision model which characterizes prior information of the multi-source data and mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
The mathematical processing module is used for carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
The decision output module is used for inputting the prior information of the mathematical multi-source data and the observation information of the multi-source data into the multi-source data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision geoscience research.
The model building module builds a multi-source data fusion decision model, which comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
The model building module obtains the decision probability by carrying out weighted average on the posterior distribution through the decision weight of prior information of the multi-source data and the decision weight of observation information of the multi-source data.
The mathematical processing module performs mathematical processing on prior information of the multi-source data, and the mathematical processing module comprises:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
According to the invention, a Bayesian statistical method is introduced in the aspect of fusion of the geoscience multisource data, the multisource data can be effectively fused, a plurality of priori information which is difficult to mathematical is converted into various probability distributions, the data utilization efficiency is effectively improved, and the decision accuracy of the prediction problem and the uncertainty problem in the geoscience research is improved.
The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application, the scope of which is defined by the claims. Various modifications and equivalent arrangements of this application will occur to those skilled in the art, and are intended to be within the spirit and scope of the application.
Claims (10)
1. The geoscience multi-source data fusion method based on Bayesian statistics is characterized by comprising the following steps of:
Determining multi-source data in a ground-based personalized teleoperation of the geoscience study, wherein the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
Acquiring prior information of multi-source data and observation information of the multi-source data;
constructing a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining prior information representing the multi-source data and a multi-source data fusion decision model of mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
and inputting the prior information of the mathematical multisource data and the observation information of the multisource data into a multisource data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision-making geoscience research.
2. The bayesian-statistics-based geoscience multi-source data fusion method according to claim 1, wherein: the construction method of the multi-source data fusion decision model comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
3. The bayesian-statistics-based geoscience multi-source data fusion method according to claim 1, wherein: and carrying out weighted average on the posterior distribution through the decision weight of the prior information of the multi-source data and the decision weight of the observation information of the multi-source data to obtain the decision probability.
4. The bayesian-statistics-based geoscience multi-source data fusion method according to claim 1, wherein: the method for determining the decision weight of the prior information of the multi-source data comprises the following steps:
Acquiring prior information of multi-source data and observation information of the multi-source data;
calculating the dispersion of prior information of the multi-source data and observation information of the multi-source data, and quantifying the instability of various data in the multi-source data by using the dispersion, wherein the dispersion is calculated by using a similarity function, and the instability is as follows:
Ki=Dis(mold,i,mnew,i);
Wherein K i is instability of ith data in the multi-source data, m old,i is priori information of the ith data in the multi-source data, m new,i is observation information of the ith data in the multi-source data, and Dis (m old,i,mnew,i) is Euclidean distance between m old,i and m new,i;
Carrying out min-max standardization processing on the instability, and linearly mapping the instability to the range of the [0,1] interval, wherein K i after the min-max standardization processing is as follows: k i'=(Ki -minK)/(maxK-minK), wherein minK is the minimum value of the instability of n data in the multi-source data, and max is the maximum value of the instability of n data in the multi-source data;
according to the instability of the multi-source data, constructing decision weights of prior information of the multi-source data which are adaptively adjusted according to the instability, wherein the decision weights of the prior information of the multi-source data are as follows:
Wi=-(1-Ki')^r*log Ki’;
Wherein W i is the decision weight of priori information of ith data in the multi-source data, K i' is the instability of the ith data in the multi-source data after min-max standardization processing, i is a counting variable, r is greater than 1, and r is the constant power.
5. The bayesian-statistics-based geoscience multi-source data fusion method according to claim 1, wherein: the method for determining the decision weight of the observation information of the multi-source data comprises the following steps:
Obtaining the decision weight of the observation information of the multi-source data according to the decision weight of the prior information of the multi-source data;
the decision weight of the observation information of the multi-source data is as follows:
Mi=1-Wi;
Wherein M i is the decision weight of the observation information of the ith data in the multi-source data, W i is the decision weight of the prior information of the ith data in the multi-source data, and i is the counting variable.
6. The bayesian-statistics-based geoscience multi-source data fusion method according to claim 1, wherein: the mathematical method for the prior information of the multi-source data comprises the following steps:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
7. A bayesian statistics-based geoscience multi-source data fusion system, characterized in that it is applied to a bayesian statistics-based geoscience multi-source data fusion method according to any one of claims 1-6, said geoscience multi-source data fusion system comprising:
The data preprocessing module is used for determining multi-source data in the ground physical teleoperation of the earth science research, and the data types of the multi-source data comprise continuous data types, classified data types and experience data types;
The data acquisition module is used for acquiring prior information of the multi-source data and observation information of the multi-source data;
The model building module is used for building a probability model which fuses the multi-source data and is used for quantitatively determining the prediction problem and the uncertainty problem in the geoscience research through a Bayesian statistical framework, and obtaining a multi-source data fusion decision model which characterizes prior information of the multi-source data and mapping relation between observation information of the multi-source data and decision probabilities of the prediction problem and the uncertainty problem in the quantitative decision geoscience research;
The mathematical processing module is used for carrying out mathematic on prior information of the multi-source data according to the data type of the multi-source data through the prior model;
The decision output module is used for inputting the prior information of the mathematical multi-source data and the observation information of the multi-source data into the multi-source data fusion decision model to obtain the decision probability of the prediction problem and the uncertainty problem in the quantitative decision geoscience research.
8. A bayesian statistics-based geoscience multi-source data fusion system according to claim 1, wherein: the model building module builds a multi-source data fusion decision model, which comprises the following steps:
The multi-source data are fused through prior distribution, the prior distribution of the multi-source data is obtained, and the prior distribution of the multi-source data is as follows: p (m) =p 1(m)p2(m)…pn (m), where p (m) is the prior distribution of the multi-source data, p 1 (m) is the prior distribution of the 1 st data in the multi-source data, p 2 (m) is the prior distribution of the 2 nd data in the multi-source data, p n (m) is the prior distribution of the n-th data in the multi-source data, and m is the identifier of the multi-source data;
Obtaining a multi-source data fusion decision model according to prior distribution of multi-source data by a Bayesian theorem formula, wherein the multi-source data fusion decision model is as follows:
p(m|d)=const*p(d|m)p(m);
Wherein p (m|d) is a posterior distribution of decision probabilities representing a prediction problem and an uncertainty problem in the geoscience study, p (d|m) is a likelihood function representing a linear or nonlinear relationship between multi-source data and observation information, p (m) is a priori distribution of the multi-source data, d is an identifier of the observation information, m is an identifier of the multi-source data, and const is a constant coefficient.
9. A bayesian statistics-based geoscience multi-source data fusion system according to claim 1, wherein: the model building module obtains the decision probability by carrying out weighted average on the posterior distribution through the decision weight of prior information of the multi-source data and the decision weight of observation information of the multi-source data.
10. A bayesian statistics-based geoscience multi-source data fusion system according to claim 1, wherein: the mathematical processing module performs mathematical processing on prior information of the multi-source data, and the mathematical processing module comprises:
when the data type of the prior information of the multi-source data is continuous data type, according to the distribution type of the prior information of the multi-source data, the prior information of the multi-source data is approximately Gaussian distribution or Cauchy distribution for mathematic;
when the data type of the prior information of the multi-source data is classified data type, adopting a Markov random field to carry out numbering classification on the categories, establishing a category random field, and carrying out mathematic on the prior information of the multi-source data through Gibbs distribution;
And when the data type of the prior information of the multi-source data is the experience information type, converting the prior information of the multi-source data into a counting matrix form, and then carrying out mathematic on the prior information of the multi-source data by utilizing a transition probability matrix of a Markov process.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410064329.8A CN117932533A (en) | 2024-01-16 | 2024-01-16 | Emotion science multi-source data fusion method and system based on Bayesian statistics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410064329.8A CN117932533A (en) | 2024-01-16 | 2024-01-16 | Emotion science multi-source data fusion method and system based on Bayesian statistics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117932533A true CN117932533A (en) | 2024-04-26 |
Family
ID=90758565
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410064329.8A Pending CN117932533A (en) | 2024-01-16 | 2024-01-16 | Emotion science multi-source data fusion method and system based on Bayesian statistics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117932533A (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104834813A (en) * | 2015-04-28 | 2015-08-12 | 南京邮电大学 | Statistic analysis processing method and device for multi-source heterogeneous data of internet of things |
CN108761574A (en) * | 2018-05-07 | 2018-11-06 | 中国电建集团北京勘测设计研究院有限公司 | Rainfall evaluation method based on Multi-source Information Fusion |
CN110059337A (en) * | 2019-02-26 | 2019-07-26 | 中国西安卫星测控中心 | Gyro system life-span prediction method based on Bayes multisource data fusion |
CN110414552A (en) * | 2019-06-14 | 2019-11-05 | 中国人民解放军海军工程大学 | A kind of spare part reliability Bayesian Assessment Method and system based on multi-source fusion |
CN113027427A (en) * | 2019-12-25 | 2021-06-25 | 中石化石油工程技术服务有限公司 | Credibility-containing formation pressure while drilling correction method based on Bayesian theory |
CN113569438A (en) * | 2021-06-04 | 2021-10-29 | 郑州大学 | Urban flood model construction method based on multi-source rainfall fusion and real-time correction |
CN116805414A (en) * | 2023-06-20 | 2023-09-26 | 福建天元通信科技有限公司 | Garbage classification method and system based on Bayesian neural network |
CN117034657A (en) * | 2023-09-12 | 2023-11-10 | 江西省地质局物化探大队 | Geophysical data processing module based on priori information constraint |
CN117370832A (en) * | 2023-09-12 | 2024-01-09 | 中国人民解放军国防科技大学 | Underwater sound target identification method and device based on Bayesian neural network |
-
2024
- 2024-01-16 CN CN202410064329.8A patent/CN117932533A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104834813A (en) * | 2015-04-28 | 2015-08-12 | 南京邮电大学 | Statistic analysis processing method and device for multi-source heterogeneous data of internet of things |
CN108761574A (en) * | 2018-05-07 | 2018-11-06 | 中国电建集团北京勘测设计研究院有限公司 | Rainfall evaluation method based on Multi-source Information Fusion |
CN110059337A (en) * | 2019-02-26 | 2019-07-26 | 中国西安卫星测控中心 | Gyro system life-span prediction method based on Bayes multisource data fusion |
CN110414552A (en) * | 2019-06-14 | 2019-11-05 | 中国人民解放军海军工程大学 | A kind of spare part reliability Bayesian Assessment Method and system based on multi-source fusion |
CN113027427A (en) * | 2019-12-25 | 2021-06-25 | 中石化石油工程技术服务有限公司 | Credibility-containing formation pressure while drilling correction method based on Bayesian theory |
CN113569438A (en) * | 2021-06-04 | 2021-10-29 | 郑州大学 | Urban flood model construction method based on multi-source rainfall fusion and real-time correction |
CN116805414A (en) * | 2023-06-20 | 2023-09-26 | 福建天元通信科技有限公司 | Garbage classification method and system based on Bayesian neural network |
CN117034657A (en) * | 2023-09-12 | 2023-11-10 | 江西省地质局物化探大队 | Geophysical data processing module based on priori information constraint |
CN117370832A (en) * | 2023-09-12 | 2024-01-09 | 中国人民解放军国防科技大学 | Underwater sound target identification method and device based on Bayesian neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109214452B (en) | HRRP target identification method based on attention depth bidirectional cyclic neural network | |
CN110929848B (en) | Training and tracking method based on multi-challenge perception learning model | |
CN112285807B (en) | Meteorological information prediction method and device | |
CN108537790B (en) | Different-source image change detection method based on coupling translation network | |
CN115237896B (en) | Data preprocessing method and system based on deep learning forecast air quality | |
CN113886372A (en) | User portrait construction method based on improved analytic hierarchy process | |
CN109685823B (en) | Target tracking method based on deep forest | |
Franchi et al. | Statistical properties of the maximum Lyapunov exponent calculated via the divergence rate method | |
CN117575858B (en) | Management method, system and storage medium of intelligent course arrangement system | |
CN112766537B (en) | Short-term electric load prediction method | |
CN117932533A (en) | Emotion science multi-source data fusion method and system based on Bayesian statistics | |
CN113095377A (en) | Dangerous driving scene data random generation method and system | |
CN117391139A (en) | Weather phenomenon prediction correction method based on improved UNet neural network | |
CN117097541A (en) | API service attack detection method, device, equipment and storage medium | |
Spezia et al. | An evolutionary Monte Carlo method for the analysis of turbidity high‐frequency time series through Markov switching autoregressive models | |
CN109657907B (en) | Quality control method and device for geographical national condition monitoring data and terminal equipment | |
CN113537313B (en) | Unbalanced data set analysis method based on WGAN training convergence | |
CN115018007A (en) | Sensitive data classification method based on improved ID3 decision tree | |
CN109272534B (en) | SAR image change detection method based on multi-granularity cascade forest model | |
Yousefi et al. | Forecasting lightpath quality of transmission and implementing uncertainty in the forecast models | |
Ivanytska et al. | Study of Methods of Complex Data Analysis that Based on Machine Learning Technologies | |
Schuhen | Order of operation for multi-stage post-processing of ensemble wind forecast trajectories | |
Lamb et al. | Reduced Order Modeling for Linearized Representations of Microphysical Process Rates | |
CN116975611B (en) | High-frequency load data generation method and system based on diffusion model ODE form | |
Shin et al. | Effective data structure for smart big data systems applying an expectation-maximization algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |