CN102496062B - Personnel information fusion method based on Spiking neural network - Google Patents

Abstract

The invention discloses a personnel information fusion method based on a Spiking neural network. The personnel information fusion method comprises the following steps of: (1) collecting information of personnel number, and storing data in a real-time database; (2) reading the data from the real-time database, preprocessing the data, mapping the data into input vectors, and storing the input vectors in a relational database in groups; (3) designing a structural model of the Spiking neural network according to the input type of a sensor; (4) reading data from the relational database, multiplying the weighted value of trust degree by data, inputting the result into the Spiking neural network, and adopting frequency coding for a sensing neural cell; (5) adopting consistency coding for a hidden neural cell; (6) adopting a Hebbian learning arithmetic, updating the sensing neural cell, updating a connecting weighted value among the sensing neural cell, the hidden neural cell and an output neural cell of the Spiking neural network model, and improving system behaviors; and (7) carrying out potentiality activation calculation on input signals of the output neural cell, and determining the number of output pulses of the output neural cell, i.e. the number of personnel in a region.

Description

A kind of fusion method of personal information based on the Spiking neural network

Technical field

The invention belongs to areas of information technology, relate in particular to the fusion method of the personal information based on the Spiking neural network of personal information collection in building, Multi-source Information Fusion, the design of Spiking neural network and modeling.

Background technology

Personnel detect and analyze is research direction emerging in building energy saving field, and also relation building safety simultaneously, and the constructing operation energy consumption is played to great impact.A large amount of constructions of intelligent building, developing rapidly of building automatic level and increasing substantially of economy and living standard, constantly promoting the demand of building to comfort level and operation energy consumption control, promoted the fast development of personnel's etection theory and technology.According to the relevant expert, analyze, Devoting Major Efforts To Developing and the personal information detection technique of applying the advanced person, raising personnel accuracy of detection, can be building energy conservation and building safety provides foundation, has very important realistic meaning.

At present, the expensive measuring instrument of the personnel with the simple analysis function on market, occurred, yet, for the Yi Zuo building, adopted the instrument of this costliness to be covered with whole building, its cost obviously is difficult to allow the people accept.For personnel in building, detect, a feasible solution is exactly to utilize advanced technology of Internet of things, wireless sensor network technology, adopt detecting sensor cheap, low-power consumption to form low-power consumption redundancy multisensor network, pass through multisource information fusion technology, raising personnel accuracy of detection, reduce testing cost, reduce significantly wiring and installation transformation to building.

Aspect multi-sensor information fusion, researchist both domestic and external has proposed multiple fusion method, as D-S evidential reasoning, method of weighted mean, Kalman filtering method, the multi-Bayes estimation technique, fuzzy logic theory, neural network, expert system etc.Yet these methods have but run into difficulty when personnel detect information in processing building.At first, the existing wireless network of personnel's detection system in building, have again wired ethernet network, and network type is various; Secondly, handled sensor type is various, can be by CO ₂, temperature, humidity, infrared, RFID, the first-class polytype sensing data of shooting form; Again, the existing discrete message of buildings inner sensor information and continuous information, some or the information of random probability distribution.Therefore, the personnel in the building detect the heterogeneous network Multi-source Information Fusion and have brought new challenge to Theory of Information Fusion and technical research.

Numerous scholars have turned to bionics to sight, and the ability of the biological treatment information in nature is very superb.Therefore, the present invention, using the Spiking neural network as the Multi-source Information Fusion instrument, proposes new method for solving personnel's test problems in building.The Spiking neural network is as third generation neural network, have front two generation neural network some advantages, but simultaneously different with second generation Sigmoid neural network from traditional first generation perceptron network again, owing to adopting the pulse firing signal as neuronic input/output signal, the Spiking neural network can closer be described actual biological nervous system, and has stronger computing power.It not only can realize the fusion of discrete message and continuous information, can also process the Information Problems of random probability distribution.Therefore, the Spiking neural network is suitable for solving a heterogeneous network Multi-source Information Fusion difficult problem very much.Therefore, the present invention not only can be building energy conservation and building safety provides foundation, and can promote the development of Spiking neural net model establishing and design, has very important realistic meaning.

Summary of the invention

The objective of the invention is, for personnel's Multi-source Information Fusion and Spiking neural net model establishing problem, to propose a kind of fusion method of personal information based on the Spiking neural network.It detects low-power consumption, redundant sensor network by the structure personnel; carry out Multi-source Information Fusion; the analytical approach that effective personnel detect data is proposed; design personnel's detection system that precision is high, cost is low; not only can guarantee to build the subregion precision, reduce operation energy consumption, reserved resource and environment; and can realize personalized control according to personnel's number, for building energy and safety management and optimal control provide foundation.

For achieving the above object, the present invention adopts following technical scheme:

A kind of fusion method of personal information based on the Spiking neural network, the steps include:

1) utilize the CO that exists in the space zone ₂Concentration sensor, infrared sensor, personnel RFID, camera and relative humidity sensor carry out the collection of personnel's information of number, and data deposit real-time data base in;

2) from reading out data real-time data base, carry out pre-service, remove unreasonable data, be mapped as input vector, grouping deposits relational database in;

3), according to sensor input type design Spiking neural network structure model, be designed to comprise the Three Tiered Network Architecture of input layer, hidden layer and output layer, the corresponding sensing neuron of one of them sensor; CO ₂The sensing neuron output signal that concentration sensor, infrared sensor and relative humidity sensor are corresponding enters hidden neuron, and sensing neuron signal corresponding to personnel RFID and camera directly enters output neuron; The output umber of pulse of the corresponding output neuron of space area people number;

4) from reading out data relational database, data multiply by degree of belief weighted value input Spiking neural network, sensing neuron proportion coding;

5) hidden neuron adopts the consistance coding, and namely hidden neuron carries out logic "and" operation by the pulse of input, and only when two pulses occurred simultaneously, hidden neuron was just in pulse of corresponding time point output;

6) adopt the Hebbian learning algorithm, the connection weights between sensing neuron, hidden neuron and the output neuron of renewal Spiking neural network model, improve system action;

7) the output neuron input signal is activated to potential and calculate, determine the output umber of pulse of output neuron, personnel's number that namely should zone.

In described step 4), described frequency coding represents with following formula:

$I_i=\mathrm{round}\left(\frac{r_i-s_i}{u_i}\right)$

In formula, I _iBe i the neuronic output umber of pulse of sensing, r _iFor sensor full range, s _iFor sensor input, u _iFor the unit pulse conversion scale, function round is bracket function.

The hidden neuron of described Spiking neural network adopts the consistance coding, and namely hidden neuron carries out logic "and" operation by the pulse of input, and only when two pulses occurred simultaneously, hidden neuron was just in pulse of corresponding time point output.

In described step 6), the Hebbian learning algorithm is as follows:

$w_{\mathrm{ij}}\left(t\right)=d\×w_{\mathrm{ij}}(t-1)+r\×\underset{t_j^{\mathrm{output}},t_i^{\mathrm{input}}}{\mathrm{\Σ}}W\left(\mathrm{\Δ}{\mathrm{\τ}}_{\mathrm{ij}}\right)$

In formula, w _IjThe connection weights between sensing neuron, hidden neuron and output neuron, $W\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\right)=\left\{\begin{array}{c}[c_1(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_1})+c_2(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_2})]\&CenterDot;\exp \left(\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{\mathrm{\&tau;}}\right),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\&GreaterEqual;0\\ {\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="HDA0000117711460000011.png"\; state="\{\{state\}\}">c</figure-callout>}}_1\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_1})+c_2\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_2}),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}<0\end{array}\right.,,$ c ₁, c ₂, τ, τ ₁And τ ₂Be constant,

D is attenuation rate, and r is learning rate, Δ τ _IjIt is i input pulse and j mistiming of exporting between pulse

In described step 7), the formula of computing activation potential is as follows:

$u_i\left(t\right)=\underset{j}{\mathrm{\&Sigma;}}\underset{t_j^{\left(l\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ij}}^g\mathrm{\&epsiv;}(t-t_j^{\left(l\right)})+\underset{k}{\mathrm{\&Sigma;}}\underset{t_k^{\left(f\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ik}}^y\mathrm{\&epsiv;}(t-t_k^{\left(l\right)})$

In formula, For the weights that are connected between sensing neuron and output neuron,

For between hidden neuron and output neuron, being connected weights, u _i(t) be the input pulse activation potential of i output neuron,

Time during l pulse of expression sensor neuron j output, Time during l pulse of expression hidden neuron k output, the ε function is

τ wherein ₃For time constant; Output neuron calculates by input signal being activated to potential, determines the output umber of pulse, personnel's number that namely should zone.

The present invention is directed to personnel's Spiking neural network structure that detected the multisensor network design in building, determined the coded system of input neuron and hidden neuron, design the connection weights and determined mode and Spiking network using Hebbian mode of learning, provided the output neuron potential computing formula of computing staff's number.At first the CO to existing in the same space zone ₂Personnel's information of number of the sensor collections such as concentration sensor, infrared sensor, personnel RFID, camera and relative humidity sensor is carried out pre-service, removes unreasonable data, is mapped as input vector; Then data after pre-service be multiply by to the sensing neuron of degree of belief weighted value input Spiking neural network input layer, carry out frequency coding; Design CO ₂Hidden neuron corresponding to sensing neuron that concentration sensor, infrared sensor and relative humidity sensor are corresponding, carry out the consistance coding; By the Hebbian learning method, determine the connection weights between sensing neuron, hidden neuron and the output neuron of Spiking neural network structure model; Finally output neuron is activated to potential and calculate, determine the output of output neuron, personnel's number that namely should zone.

Beneficial effect of the present invention: the present invention adopts third generation neural network Spiking neural network, the mode of simulation biological treatment information is carried out Multi-source Information Fusion, thereby obtain that precision is higher, adaptability is stronger, the more excellent personnel's number fusion results of robustness, for Environmental security and building energy conservation provide more reliable foundation.

The accompanying drawing explanation

Fig. 1 is the Spiking neural network structure model of personal information fusion method;

Fig. 2 is a kind of fusion method of personal information based on Spiking neural network process flow diagram.

Embodiment

Below in conjunction with accompanying drawing and embodiment, a kind of fusion method of personal information based on the Spiking neural network of the present invention is described in detail.

The present invention proposes a kind of fusion method of personal information based on the Spiking neural network, is applied to build interior personnel's Multi-source Information Fusion and statistics.For the CO that exists in area of space ₂Personnel's information of number of concentration sensor, infrared sensor, personnel RFID, camera and five kinds of sensor collections of relative humidity sensor is carried out pre-service, removes unreasonable data, is mapped as input vector; According to sensor input type design Spiking neural network structure model, the network structure of design comprises input layer, sensing neuron layer, hidden neuron layer, output neuron layer, output layer, as shown in Figure 1.Wherein, the corresponding sensing neuron of sensor, CO ₂The sensing neuron output signal that concentration sensor, infrared sensor and relative humidity sensor are corresponding enters hidden neuron, and input neuron corresponding to personnel RFID and camera directly enters output neuron.To multiply by through pretreated data the sensing neuron of degree of belief weighted value input Spiking neural network, sensing neuron sample frequency coding, so in Fig. 1 ₁～so _lFor CO ₂The pretreated data of concentration sensor, l is CO ₂The number of concentration sensor; Sr ₁～sr _mFor the pretreated data of infrared sensor, m is CO ₂The number of concentration sensor; Sh ₁～sh _nFor the pretreated data of humidity sensor, n is CO ₂The number of concentration sensor; Sd ₁～sd _jFor the pretreated data of RFID sensor, j is the number of RFID sensor; Sa ₁～sa _kFor the pretreated data of ccd sensor, k is the number of ccd sensor; p _sumFor the output umber of pulse of output neuron, i.e. personnel's number in this area of space.

Accompanying drawing 2 is based on the process flow diagram of the personal information fusion method of Spiking neural network.At first the various data of sensor collection enter real-time data base, then from real-time data base, reading, carry out pre-service, and grouping deposits relational database in.By sensor type, carry out reading, process respectively CO ₂The sensing neuron output signal that concentration sensor, infrared sensor and relative humidity sensor are corresponding enters hidden neuron.Output neuron receives hidden neuron and the neuronic pulse of sensing, activates potential and calculates, and then by the output umber of pulse, determines the personnel's data in this construction area.The Spiking neural network utilizes the input pulse of output neuron and output pulse to carry out Hebbian study, upgrades the connection weights between hidden neuron and output neuron, improves system action.

The sensing neuron proportion of Spiking neural network coding in the present invention represents with following formula:

$I_i=\mathrm{round}\left(\frac{r_i-s_i}{u_i}\right)$

In formula, I _iBe i the neuronic output umber of pulse of sensing, r _iFor sensor full range, s _iFor sensor input, u _iFor the unit pulse conversion scale, function round is bracket function.

CO ₂The sensing neuron output signal that concentration sensor, infrared sensor and relative humidity sensor are corresponding enters hidden neuron, and hidden neuron adopts the consistance coding.Hidden neuron carries out logic "and" operation by the pulse of input, and only when two pulses occurred simultaneously, hidden neuron was just in pulse of corresponding time point output.

The formula of the output neuron computing activation potential of Spiking neural network is as follows:

$u_i\left(t\right)=\underset{j}{\mathrm{\&Sigma;}}\underset{t_j^{\left(l\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ij}}^g\mathrm{\&epsiv;}(t-t_j^{\left(l\right)})+\underset{k}{\mathrm{\&Sigma;}}\underset{t_k^{\left(f\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ik}}^y\mathrm{\&epsiv;}(t-t_k^{\left(l\right)})$

In formula,

For the weights that are connected between sensing neuron and output neuron,

For between hidden neuron and output neuron, being connected weights, u _i(t) be the input pulse activation potential of i output neuron, Time during l pulse of expression sensor neuron j output,

Time during l pulse of expression hidden neuron k output, the ε function is

τ wherein ₃For time constant; Output neuron calculates by input signal being activated to potential, determines the output umber of pulse, personnel's number that namely should zone.

Spiking neural network in the present invention adopts the Hebbian learning algorithm, and expression is as follows:

$w_{\mathrm{ij}}\left(t\right)=d\&times;w_{\mathrm{ij}}(t-1)+r\&times;\underset{t_j^{\mathrm{output}},t_i^{\mathrm{input}}}{\mathrm{\&Sigma;}}W\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\right)$

In formula, w _IjThe connection weights between sensing neuron, hidden neuron and output neuron, $W\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\right)=\left\{\begin{array}{c}[c_1(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_1})+c_2(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_2})]\&CenterDot;\exp \left(\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{\mathrm{\&tau;}}\right),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\&GreaterEqual;0\\ {\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="HDA0000117711460000011.png"\; state="\{\{state\}\}">c</figure-callout>}}_1\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_1})+c_2\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_2}),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}<0\end{array}\right.,,$ c ₁, c ₂, τ, τ ₁And τ ₂Be constant, D is attenuation rate, and r is learning rate, Δ τ _IjIt is i input pulse and j mistiming of exporting between pulse

Although principle of the present invention is showed in conjunction with the embodiments and is described, but it will be understood to those of skill in the art that in the situation that do not depart from principle of the present invention and essence, change these embodiments, as every layer of neuron number of Spiking neural network model, the change of input layer coding form etc., its scope also falls in claim of the present invention and equivalent limited range thereof.

Claims (4)

1. the fusion method of the personal information based on the Spiking neural network, is characterized in that,

1) utilize the CO that exists in the space zone ₂Concentration sensor, infrared sensor, personnel RFID, camera and relative humidity sensor carry out the collection of personnel's information of number, and data deposit real-time data base in;

2) from reading out data real-time data base, carry out pre-service, remove unreasonable data, be mapped as input vector, grouping deposits relational database in;

3), according to sensor input type design Spiking neural network structure model, be designed to comprise the Three Tiered Network Architecture of input layer, hidden layer and output layer, the corresponding sensing neuron of one of them sensor; CO ₂The sensing neuron output signal that concentration sensor, infrared sensor and relative humidity sensor are corresponding enters hidden neuron, and sensing neuron signal corresponding to personnel RFID and camera directly enters output neuron; The output umber of pulse of the corresponding output neuron of space area people number;

4) from reading out data relational database, data multiply by degree of belief weighted value input Spiking neural network, sensing neuron proportion coding;

5) hidden neuron adopts the consistance coding, and namely hidden neuron carries out logic "and" operation by the pulse of input, and only when two pulses occurred simultaneously, hidden neuron was just in pulse of corresponding time point output;

6) adopt the Hebbian learning algorithm, the connection weights between sensing neuron, hidden neuron and the output neuron of renewal Spiking neural network model, improve system action;

7) the output neuron input signal is activated to potential and calculate, determine the output umber of pulse of output neuron, personnel's number that namely should zone.

2. the fusion method of the personal information based on the Spiking neural network as claimed in claim 1, is characterized in that, in described step 4), described frequency coding represents with following formula:

$I_i=\mathrm{round}\left(\frac{r_i-s_i}{u_i}\right)$

, in formula, I _iBe i the neuronic output umber of pulse of sensing, r _iFor sensor full range, s _iFor sensor input, u _iFor the unit pulse conversion scale, function round is bracket function.

3. the fusion method of the personal information based on the Spiking neural network as claimed in claim 1, is characterized in that, in described step 6), the Hebbian learning algorithm is as follows: algorithm is as follows:

$w_{\mathrm{ij}}\left(t\right)=d\&times;w_{\mathrm{ij}}(t-1)+r\&times;\underset{t_j^{\mathrm{output}},t_i^{\mathrm{input}}}{\mathrm{\&Sigma;}}W\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\right)$

In formula, w _IjThe connection weights between sensing neuron, hidden neuron and output neuron, $W\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\right)=\left\{\begin{array}{c}[c_1(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_1})+c_2(1-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\widetilde{\mathrm{\&tau;}}}_2})]\&CenterDot;\exp \left(\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{\mathrm{\&tau;}}\right),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\&GreaterEqual;0\\ c_1\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_1})+c_2\exp (-\frac{\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}}{{\mathrm{\&tau;}}_2}),\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}<0\end{array}\right.,$ c ₁, c ₂, τ, τ ₁And τ ₂Be constant,

D is attenuation rate, and r is learning rate, Δ τ _IjIt is i input pulse and j mistiming of exporting between pulse

4. the fusion method of the personal information based on the Spiking neural network as claimed in claim 1, is characterized in that, in described step 7), the formula of computing activation potential is as follows:

$u_i\left(t\right)=\underset{j}{\mathrm{\&Sigma;}}\underset{t_j^{\left(l\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ij}}^g\mathrm{\&epsiv;}(t-t_j^{\left(l\right)})+\underset{k}{\mathrm{\&Sigma;}}\underset{t_k^{\left(f\right)}}{\mathrm{\&Sigma;}}{w}_{\mathrm{ik}}^y\mathrm{\&epsiv;}(t-t_k^{\left(l\right)})$

In formula,

For the weights that are connected between sensing neuron and output neuron,

For between hidden neuron and output neuron, being connected weights, u _i(t) be the input pulse activation potential of i output neuron,

Time during l pulse of expression sensor neuron j output,

Time during l pulse of expression hidden neuron k output, the ε function is τ wherein ₃For time constant; Output neuron calculates by input signal being activated to potential, determines the output umber of pulse, personnel's number that namely should zone.

Images