CN102759430B - BP (Back Propagation) neural network based high-precision correction and test method for resonance cylinder pressure sensor - Google Patents

Abstract

The invention belongs to the technical field of test and measurement and particularly relates to a BP (Back Propagation) neural network based high-precision correction and test method for a resonance cylinder pressure sensor, aiming to increase the correction and test precision for the resonance cylinder pressure sensor. The method comprises the steps of: structuring a sensor BP neural network of a dual-implicit-strata network structure, wherein input variables of the network structure are respectively the output period T and the temperature voltage V of the sensor, and output variables of the network structure is the pressure value P; acquiring output periods and temperature voltages of the sensor at different temperature under different pressure input conditions; and collecting output quantities of the sensor at different temperature under different pressure conditions to serve as correction and test as well as inspection sample points. According to the method, with the dual-implicit-strata network structure, the number of network parameters is reduced while the output precision is guaranteed; and the correction and test precision of the resonance cylinder pressure sensor is increased by 25%, and the high-precision correction and test of the sensor is realized.

Description

Resonant vibration barrel pressure pickup high precision school method for testing based on BP neural network

Technical field

The invention belongs to a kind of thermometrically technical field, relate to resonant vibration barrel pressure pickup high precision school method for testing, be specifically related to a kind of resonant vibration barrel pressure pickup high precision school method for testing based on BP neural network.

Background technology

For realizing the accurate control of guided missile, day by day improve for tonometric range and accuracy requirement, therefore need to develop high-precision pressure sensor.For current good stability, resonant vibration barrel pressure pickup that precision is high, the level of high precision school examination technology becomes the key factor that affects sensor accuracy, the polynomial fitting method based on sensor physical model in the past conventionally adopting, vibration cylinder pressure transducer output periodic signal T and temperature voltage signal V, according to sensor physical model, between they and input measurement pressure P, can set up following nonlinear multivariate regression equations:

$P-P_0=\underset{i=0}{\overset{m}{\mathrm{\Σ}}}\underset{j=0}{\overset{n}{\mathrm{\Σ}}}{K}_{\mathrm{ij}}{(T-T_0)}^i{(V-V_0)}^j$ Formula (1)

In formula: P---input measurement air pressure;

T---the sensor output cycle;

V---baroceptor temperature voltage;

P ₀---reference point air pressure;

T ₀---the reference point sensor output cycle;

V ₀---reference point temperature voltage;

K _ij---the characteristic coefficient of sensor.

Can there is a different set of K for each vibration cylinder pressure transducer _ijsystem of parameters.

Traditional Polynomial Fitting Technique is on the basis of formula (1), selects different polynomial expression orders, utilizes least square method to obtain K _ijcoefficient.The method can not meet the accuracy requirement of model 0.028%FS on fitting precision.

Summary of the invention

The object of the invention is to improve resonant vibration barrel pressure pickup school examination precision, a kind of resonant vibration barrel pressure pickup high precision school method for testing based on BP neural network is provided.

The technical solution adopted in the present invention is:

A resonant vibration barrel pressure pickup school method for testing based on BP neural network, comprises the sensor BP neural network that builds two hidden layer network structures, the output cycle T that the input variable that makes network structure is sensor and temperature voltage V, and output variable is pressure value P; Output cycle and the temperature voltage of pick-up transducers under different temperatures, different pressures initial conditions; Gather different temperatures and pressure condition lower sensor output quantity as school examination and test samples point.

A kind of resonant vibration barrel pressure pickup school method for testing based on BP neural network as above, wherein: in described sensor BP neural network, utilize tanh S type transport function as the transport function between input layer and first middle layer, first middle layer and second middle layer, utilize linear transfer function as the transport function between second middle layer and output layer.

A kind of resonant vibration barrel pressure pickup school method for testing based on BP neural network as above, wherein: when described sample point gathers, within the scope of-45 DEG C to 80 DEG C, choose multiple temperature spots, each temperature spot is chosen multiple sampled points.

A kind of resonant vibration barrel pressure pickup school method for testing based on BP neural network as above, wherein: in described sensor BP neural network, the network weight of input layer and first middle layer node is k-factor, the network weight of first middle layer node and second middle layer node is W coefficient, and the network weight of second middle layer node and output layer node is s coefficient; And the nodes in each middle layer is 4; And tanh S type transport function is as follows:

$y_i=\frac{2}{1+e^{-2(\underset{j=1}{\overset{2}{\mathrm{\Σ}}}{k}_{\mathrm{ij}}{x}_j-{\mathrm{\θ}}_i)}}-1(i=\mathrm{1,2}...4;j=\mathrm{1,2})$

$Y_m=\frac{2}{1+e^{-2(\underset{i=1}{\overset{4}{\mathrm{\Σ}}}{W}_{\mathrm{mi}}{y}_i-{\mathrm{\ω}}_m)}}-1(m=\mathrm{1,2}...4;i=\mathrm{1,2}...4)$

Wherein: y _ifor the output in first middle layer; x _jfor input variable value; k _ijfor the network weight in input layer and first middle layer; θ _ifor the threshold value of first each node in middle layer;

Y _mbe the output in second middle layer; W _mifor the network weight in first middle layer and second middle layer; ω _mit is the threshold value of second each node in middle layer;

Linear transfer function is as follows:

$O=\underset{n=1}{\overset{4}{\mathrm{\Σ}}}{s}_{\ln }{Y}_n-{\mathrm{\β}}_l(l=1;n=\mathrm{1,2}...4)$

Wherein: O is model output; s _lnit is the network weight of second middle layer and output layer; β _lfor the threshold value of output layer node.

A kind of resonant vibration barrel pressure pickup school method for testing based on BP neural network as above, wherein: adopt the training method of fast algorithm as BP neural network, threshold value and weights to each layer in training are revised, and error function is declined along negative gradient direction.

The invention has the beneficial effects as follows:

School provided by the invention method for testing has used two hidden layer network structures, in ensureing output accuracy, has reduced network parameter quantity; Adopt the method for ground training, effectively controlled the reliability of training process; By solidifying initialization training parameter, solve the problem reproduced of training process; And built S type fitting function, and strengthen the approximation capability of matched curve, improve fitting precision compared with fitting of a polynomial.

The present invention, by building BP neural network model, has improved 25% by resonant vibration barrel pressure pickup school examination precision, has solved sensor high precision school and may well ask and topic can meet guided missile accurate pressure measurement demand.

Brief description of the drawings

Fig. 1 is the sensor BP network model figure that the present invention adopts;

Fig. 2 is pressure error of fitting comparison diagram.

Embodiment

Below in conjunction with drawings and Examples, a kind of resonant vibration barrel pressure pickup high precision school method for testing based on BP neural network provided by the invention is introduced:

A resonant vibration barrel pressure pickup school method for testing based on BP neural network, comprises the sensor BP neural network that builds two hidden layer network structures, the output cycle T that the input variable that makes network structure is sensor and temperature voltage V, and output variable is pressure value P; Output cycle and the temperature voltage of pick-up transducers under different temperatures, different pressures initial conditions; Gather different temperatures and pressure condition lower sensor output quantity as school examination and test samples point.

In sensor BP neural network, utilize tanh S type transport function as the transport function between input layer and first middle layer, first middle layer and second middle layer, utilize linear transfer function as the transport function between second middle layer and output layer.

When sample point gathers, within the scope of-45 DEG C to 80 DEG C, choose multiple temperature spots, each temperature spot is chosen multiple sampled points.

The present invention adopts BP neural network to build resonant vibration barrel pressure pickup school die trial type, and particular content is as follows:

First carry out sample data collection, gather output cycle and the temperature voltage of pressure transducer under different temperatures, different pressures initial conditions.For example test temperature point is chosen 14 temperature spots within the scope of-45 DEG C to 80 DEG C, under each temperature spot, pressure measurement point is chosen respectively pressure from high to low and is amounted to 35 spot pressures, after sensor stable output signal, record test figure, make wherein a part of for sample point is tried in school, a part is test samples point.The sample number gathering should be tried one's best many.

In sensor BP neural network builds, select two hidden layer network structures, the output cycle T that the input variable that makes network structure is sensor and temperature voltage V, output variable is pressure value P; Variable T, V, P are variable X 1, X2, O through normalized, and BP network model figure as shown in Figure 1.

Owing to determining that the number of plies in middle layer and the nodes in each middle layer are very complicated problems always, there is no theoretic guidance, often need rule of thumb to determine with many experiments.The present invention is by data simulation repeatedly, determine the middle layer number of plies be preferably 2 and the nodes in each middle layer be preferably at 4 o'clock (for example, 2,4,4,1 structures), fitting effect is better, also can adopt other numerical value.

In Fig. 1, the network weight of input layer and first middle layer node is k-factor, and the network weight of first middle layer node and second middle layer node is W coefficient, and the network weight of second middle layer node and output layer node is s coefficient.Network model utilizes tanh S type transport function as the transport function between input layer and first middle layer, first middle layer and second middle layer, and formula (2) and formula (3) are shown in function design:

$y_i=\frac{2}{1+e^{-2(\underset{j=1}{\overset{2}{\mathrm{\Σ}}}{k}_{\mathrm{ij}}{x}_j-{\mathrm{\θ}}_i)}}-1(i=\mathrm{1,2}...4;j=\mathrm{1,2})$ Formula (2)

$Y_m=\frac{2}{1+e^{-2(\underset{i=1}{\overset{4}{\mathrm{\Σ}}}{W}_{\mathrm{mi}}{y}_i-{\mathrm{\ω}}_m)}}-1(m=\mathrm{1,2}...4;i=\mathrm{1,2}...4)$ Formula (3)

Wherein: y _ifor the output in first middle layer; x _jfor input variable value; k _ijfor the network weight in input layer and first middle layer; θ _ifor the threshold value of first each node in middle layer.Y _mbe the output in second middle layer; W _mifor the network weight in first middle layer and second middle layer; ω _mit is the threshold value of second each node in middle layer.

Choose linear transfer function as the transport function between second middle layer and output layer, formula (4) is shown in function design:

$O=\underset{n=1}{\overset{4}{\mathrm{\Σ}}}{s}_{\ln }{Y}_n-{\mathrm{\β}}_l(l=1;n=\mathrm{1,2}...4)$ Formula (4)

Wherein: O is model output; Can obtain output pressure P through renormalization processing; s _lnit is the network weight of second middle layer and output layer; β _lfor the threshold value of output layer node.

Utilize sample data to carry out model training based on neural network tool platform, threshold value and weights to each layer in training are revised, and error function is declined along negative gradient direction.The value that obtains network parameter through sample data training, wherein certain training result is as follows:

k _ij=

-3.19283709 -7.16055966

3.07537706 -0.04163307

6.06282238 -9.60802625

-0.99771530 6.42799254

θ _i=

12.46850700

-3.48749525

-2.23594779

-1.30846304

W _mi=

0.06106101 -2.72750488 0.00198993 0.16069464

0.03575073 1.17094160 0.02075128 0.08536263

-0.23359155 0.61221741 1.11657418 0.15811123

0.41401469 0.41214660 1.96806393 0.08391720

ω _m=

2.75440664

1.79626179

-0.94786308

1.49824852

s _ln=

-0.02919945 0.06017406 0.05557906 0.00022849

β _l=

0.88376202

Preferably adopt fast algorithm (LM algorithm) as the training method of BP neural network and progressively by parameter adjustments such as e-learning rate, momentum term, iterations and targets to optimum condition.Utilize sample data to carry out model training based on MATLAB network tool platform, threshold value and weights to each layer in training are revised, and error function is declined along negative gradient direction.Obtain the value of each network parameter through sample data training.

In order to compare the precision of BP neural network model of fit and fitting of a polynomial model, calculate 2 times of residual standard deviation 2S of test sample _p, fitting of a polynomial model is 0.250kPa, and BP neural network model of fit is 0.188kPa, and fitting precision has improved approximately 25%, as shown in Figure 2.Fig. 2 is different temperatures environment downforce measuring error contrast effect figure, solid line representative polynomial matching pressure measurement errors, and dotted line represents neural network model measuring error.

Artificial intelligence neural networks is applied to the examination of sensor school by the present invention, selects BP neural network to build resonant vibration barrel pressure pickup school die trial type, realizes the output of sensor high-precision pressure, meets the accurate demand for control of guided missile.

Claims (2)

1. the resonant vibration barrel pressure pickup school method for testing based on BP neural network, comprises the sensor BP neural network that builds two hidden layer network structures, the output cycle T that the input variable that makes network structure is sensor and temperature voltage V, and output variable is pressure value P; Output cycle and the temperature voltage of pick-up transducers under different temperatures, different pressures initial conditions; Gather different temperatures and pressure condition lower sensor output quantity as school examination and test samples point;

In described sensor BP neural network, utilize tanh S type transport function as the transport function between input layer and first middle layer, first middle layer and second middle layer, utilize linear transfer function as the transport function between second middle layer and output layer;

When described sample point gathers, within the scope of-45 DEG C to 80 DEG C, choose multiple temperature spots, each temperature spot is chosen multiple sampled points;

In described sensor BP neural network, the network weight of input layer and first middle layer node is k-factor, the network weight of first middle layer node and second middle layer node is W coefficient, and the network weight of second middle layer node and output layer node is s coefficient; And the nodes in each middle layer is 4; And tanh S type transport function is as follows:

$y_i=\frac{2}{{1+e}^{-2(\underset{j=1}{\overset{2}{\mathrm{\Σ}}}{k}_{\mathrm{ij}}{x}_j-{\mathrm{\θ}}_i)}}-1(i=\mathrm{1,2}...4;j=\mathrm{1,2})$

$Y_m=\frac{2}{1+e^{-2(\underset{i=1}{\overset{4}{\mathrm{\Σ}}}{W}_{\mathrm{mi}}{y}_i-{\mathrm{\ω}}_m)}}-1(m=\mathrm{1,2}...4;i=\mathrm{1,2}...4)$

Wherein: y _ifor the output in first middle layer; x _jfor input variable value; k _ijfor the network weight in input layer and first middle layer; θ _ifor the threshold value of first each node in middle layer;

Y _mbe the output in second middle layer; W _mifor the network weight in first middle layer and second middle layer; ω _mit is the threshold value of second each node in middle layer;

Linear transfer function is as follows:

$O=\underset{n=1}{\overset{4}{\mathrm{\Σ}}}{s}_{\ln }{Y}_n-{\mathrm{\β}}_l(l=1;n=\mathrm{1,2}...4)$

Wherein: O is model output; s _lnit is the network weight of second middle layer and output layer; β _lfor the threshold value of output layer node.

2. a kind of resonant vibration barrel pressure pickup school method for testing based on BP neural network according to claim 1, it is characterized in that: adopt the training method of fast algorithm as BP neural network, threshold value and weights to each layer in training are revised, and error function is declined along negative gradient direction.