CN102360180A - Method for identifying brake performance of safety monitoring system for motor vehicles - Google Patents

Abstract

The invention discloses a method for identifying the brake performance of a safety monitoring system for motor vehicles. The method comprises the following steps: obtaining operating variables of acceleration, pressure and temperature input and brake performance parameter output; analyzing a redundant information relation between the operating variables by using a principal component analysis method so as to eliminate redundant data; and determining the structure of a recurrent wavelet ELman neural network, and obtaining a plurality of accurate values of brake performance parameters for the operating process by using an identification method. In the invention, through carrying out identification processing on obtained brake performance sensing quantities, accurate sensing information can be provided for a rear-end safety status forecasting module; the method can be used for carrying out identification on operating multi-parameter input information of motor vehicles, therefore, the method has strong learning and generalization capabilities; and by using the method disclosed by the invention, inevitable error accumulation caused by a specific computation model can be shielded, therefore, the accuracy of safe-operating brake performance monitoring parameters of motor vehicles can be effectively improved.

Description

Motor vehicle safety monitoring system braking ability discrimination method

Technical field

The present invention relates to motor vehicle safe operation parameter monitoring, relate in particular to a kind of information processing method of the motor vehicle safe operation monitoring system braking ability identification based on recurrence small echo Elman network.

Background technology

Motor vehicle security of operation Condition Monitoring Technology is the main means that guarantee motor vehicle safe drive, also is the inexorable trend of motor vehicle security of operation detection technique development.Adopt motor vehicle security of operation condition monitoring technology that motor vehicle security of operation state and operating index are carried out dynamic monitoring; In time find and the prevention vehicle trouble; Development is monitored, controls, manages and made a strategic decision in the safety supervisory network system of one, and is significant to the motor vehicle safe operation.It can not only improve technical guarantee ability, the minimizing traffic hazard of motor vehicle safe operation, and the development that promotes automotive industry and communications and transportation cause is had significant meaning.Motor vehicle security of operation status monitoring mainly comprises monitoring motor vehicle (vehicle body, wheel) athletic posture parameter, dynamic loading parameter, braking ability parameter.Braking ability is to estimate the most important technical indicator of motor vehicle, is one of elementary item of automotive safety detection.Motor vehicle reruns in the process, can produce braking, quicken, turn to, operating mode such as straight-line travelling, through monitoring wheel and vehicle performance can obtain more directly, truer, more rich motor vehicle safe operation information.

The gyro free strap down measuring system that motor vehicle running status braking ability monitoring is adopted be a kind of multivariate, non-linear, the time complication system that becomes; When considering alignment error; Sensitive direction matrix and the installation site matrix that is installed on 12 accelerometers of vehicle body all can have subtle change on itself vector; Thereby angular velocity that calculates and theoretical value have than large deviation, and this deviation can constantly be accumulated along with the time.Influence the influence that the factor of system accuracy is resolved except the alignment error angle speed, be still waiting to solve for other factor that influences system accuracy, but because the model of setting up is complicated, and calculated amount is loaded down with trivial details, it is thorough to be difficult to consider.If need not know concrete precise math model, can consider so to carry out System Discrimination with neural network method, approach non-linear between system's input and output.

Wavelet transformation is a kind of signal processing method that on the Fourier analysis basis, grows up; Wavelet transformation can resolve into the block signal on the different frequency bands with the different frequency mixed signal of various weave ins; Observe respectively and handle in each time-frequency region then, all have the ability of characterization signal local feature in time domain and frequency domain.In view of wavelet transformation has good time-frequency local property and zoom character, and the Elman network has self study, strong robustness and popularization ability, just can combine both advantages wavelet transformation and Elman network being combined modeling, improves the accuracy of modeling.Therefore the present invention sets forth a kind of motor vehicle safe operation monitoring system braking ability discrimination method based on recurrence small echo Elman network.

Severity of braking parameter when " the improved motor vehicle braking system " of domestic (patent of invention) number CN1605502 brakes through motor vehicle is run at high speed is handled classification; And take the graded signal mode to show; Clear and definite car brakeing intensity; So that the back car of running at high speed is taked to brake counter-measure accordingly timely and accurately, avoid the back car blindly to brake the danger that causes.This patent does not relate to the signal processing method of motor vehicle braking ability discrimination.

Domestic (patent of invention) number CN102112351A " brake system in the motor vehicle being carried out the method for functional check " relates to a kind of method that is used for the brake system of motor vehicle is carried out functional check; At first brake fluid is moved in the apotheca in the method; Measure the brake pressure in the brake circuit subsequently and this brake pressure compared with reference value, wherein produce alerting signal being lower than under the situation of reference value.This patent does not relate to the signal processing method of motor vehicle braking ability identification.

Domestic (patent of invention) number CN101480946 that proposed in 2008 " a kind of based on the wheel-loaded intelligent sensing wheel brake performance monitoring methods " discloses a kind of based on the wheel-loaded intelligent sensing wheel brake performance monitoring methods; Through the intelligent sensing module being installed on the wheel hub equatorial plane of each wheel of motor vehicle and at vehicle body the intelligent sensing unit being installed, transducing signal obtains the main braking ability parameter of wheel through signal condition, digitizing, attitude algorithm, braking algorithm computation.This method merges and analysis through parameter being carried out many sensing datas, can initiatively estimate motor vehicle sports safety situation.This patent content relates to based on braking accurate model mode and carries out the braking ability calculation of parameter, the identification signal New Method for Processing based on the non-concrete model of recurrence small echo Elman network that does not relate to that this patent proposes.

Domestic (patent of invention) number CN101718769A " a kind of source drive-type 235U concentration recognition method based on parallel type genetic Elman neural network " comprises mainly and sets up neural network model that the neural network model structure is divided into three layers: data allocations layer, subnet layer, decision making package layer; Neutron pulse signal auto-correlation function to obtaining is done pre-service; With the data allocations layer of the input of the signal auto-correlation function sample after handling parallel type genetic Elman network, adopt the multiple spot sampling that circulates at random that sample data is distributed; Each hereditary Elman subnet that the data of distributing are imported respectively in the subnet layer is discerned, and provided recognition result separately; The decision making package layer is made joint disposal through the recognition result to a plurality of subnets, draws the final recognition result of 235U concentration.This method is because of its higher data utilization factor, and novel network structure has obtained 235U concentration recognition effect preferably.This patent does not relate to the signal processing method of recurrence small echo Elman network.

Summary of the invention

For solving, the invention provides a kind of motor vehicle safety monitoring system braking ability discrimination method because of factor insoluble error problems that parameter measurement is brought to braking ability such as installation, sensing demarcation.Non-concrete model identification disposal route based on recurrence small echo Elman network can effectively improve the accuracy of braking ability measuring multiple parameters value.Said technical scheme is following:

Motor vehicle safety monitoring system braking ability discrimination method comprises:

A obtains acceleration, pressure, temperature input and braking ability parameter output function variable;

B eliminates data redundancy through redundant information relation between the PCA analysis operation variable;

C confirms recurrence small echo ELman neural network structure, and obtains a plurality of braking ability of operational process accurately parameter values through discrimination method.

The beneficial effect of technical scheme provided by the invention is:

The recurrence small echo Elman network discrimination method of setting up according to this method has generality; And also exist similar problem in other parameter of motor vehicle safe operation; Therefore this method can also be generalized to other parameter detecting of motor vehicle operation, like aspects such as attitude, dynamic loadings.

Description of drawings

Fig. 1 is based on recurrence small echo Elman network machine motor-car safety monitoring system braking ability discrimination method modeling process flow diagram;

Fig. 2 is based on the hereditary training algorithm process flow diagram in the recurrence small echo Elman network machine motor-car safety monitoring system braking ability discrimination method;

Fig. 3 is a recurrence small echo Elman schematic network structure;

Fig. 4 a, 4b and 4c are based on the motor vehicle safety monitoring system braking ability discrimination method of recurrence small echo Elman network and implement platform illustration in kind.

Embodiment

For making the object of the invention, technical scheme and advantage clearer, will combine accompanying drawing that embodiment of the present invention is done to describe in detail further below:

Present embodiment provides a kind of method of motor vehicle safety monitoring system braking ability discrimination method.

Referring to Fig. 1, this method is based on that recurrence small echo Elman realizes, may further comprise the steps:

Step 101 obtains acceleration, pressure, temperature input and braking ability parameter output function variable;

Step 102 is eliminated data redundancy through redundant information relation between the PCA analysis operation variable.

Step 103 input data normalization is handled;

Step 104 is utilized experimental formula and experiment algorithm computation network model hidden layer node number;

For being had, model better approaches performance and generalization ability; Adopt experimental formula to choose the method that combines and confirm choosing of hidden layer neuron number with experiment; Be input number of nodes at first according to experimental formula

; N is the output node number; L is the constant between the 1-20) confirm the scope of the number of a hidden layer neuron; Set corresponding training step number and training precision, optimum hidden layer neuron number is therefrom chosen at last in preference pattern training from small to large then.

Step 105 recurrence small echo Elman neural network model parameter initialization:

The initial setting up of

The initial setting up of

specifically be divided into four the step realize:

1. produce the initial setting up value of equally distributed random number as

on [1,1] interval at first at random;

2. then

carried out normalization by row:

${\mathrm{<figure-callout\; id="2"\; label="W\; ij"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ij}}^{}=\frac{{\mathrm{<figure-callout\; id="2"\; label="W\; ij"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ij}}^{}}{\sqrt{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(W_{\mathrm{ij}}^2\right)}^2}},i=1,2,\&CenterDot;\&CenterDot;\&CenterDot;,m---\left(1\right)$

3. then multiply by one again and count n with input layer, hidden layer neuron is counted the relevant factor of m and transport function:

${\mathrm{<figure-callout\; id="2"\; label="W\; ij"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ij}}^{}=C*n^{\frac{1}{m}}*{\mathrm{<figure-callout\; id="2"\; label="W\; ij"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ij}}^{},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,m---\left(2\right)$

C is a constant relevant with the hidden layer transport function in the formula.The value of C is extremely important to network, through study practice repeatedly relatively, to the proper value of Morlet small echo Elman neural network between 2.3～2.6;

4. last and training sample is got in touch.If maximal value is x in input layer j the neuronic input sample _Jmax, minimum value is x _Jmin, then:

$W_{\mathrm{<figure-callout\; id="2"\; label="ij"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">ij</figure-callout>}}^2=\frac{2*W_{\mathrm{ij}}^2}{x_{j\max }-x_{j\min }},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,m---\left(3\right)$

that obtain by above step is the initial weight of input layer to hidden layer.

The initial setting up of

Obtain after ; And then carry out the initialization of of hidden layer neuron threshold value, step is following:

1. produce the initial setting up value of equally distributed random number as

on [1,1] interval at first at random;

2. then multiply by an input layer again and count n, hidden layer neuron is counted the m and the relevant factor of transport function;

$W_{i0}^2=C*n^{\frac{1}{m}}*W_{i0}^2,i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,m---\left(4\right)$

3. at last, training sample gets in touch with reaching

.

$W_{i0}^2=W_{i0}^2-0.5*\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2(x_{j\max }+x_{j\min }),j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,m---\left(5\right)$

(3) the flexible initial setting up with translation parameters of small echo

After being provided with initial weight, it also is very important that the flexible translation parameters of small echo is carried out initial setting up, generally is divided into two kinds of situation.

1. the input layer number is 1.Then to the neuronic flexible parameter a of each small echo _iGet identical value, and translation parameters b _i=(i-1) L/m, i=1,2 ..., m.L is a number of training in the formula, and m is the neuron number of transport function place layer.

2. the node number of input layer is greater than 1.Know by the small echo basic theory, if the time domain center of female small echo is t ^*, radius is a Δ _ΨThen the flexible concentrated area that ties up to time domain of small echo is:

[b+at ^*-aΔ _Ψ，b+at ^*+aΔ _Ψ]

In order to make the flexible system of small echo cover the gamut of input vector, then the initial value setting of flexible translation parameters must be satisfied following formula:

$b_i+a_i{t}^{*}-a_i{\mathrm{\&Delta;}}_{\mathrm{\&psi;}}=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\min }---\left(6\right)$

$b_i+a_i{t}^{*}+a_i{\mathrm{\&Delta;}}_{\mathrm{\&psi;}}=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\max }---\left(7\right)$

Can obtain by following formula:

$a_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\max }-\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\mathrm{<figure-callout\; id="2"\; label="min"\; filenames="HSA00000584101000032.png"\; state="\{\{state\}\}">min</figure-callout>}}}{2{\mathrm{\&Delta;}}_{\mathrm{\&psi;}}}---\left(8\right)$

$b_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\max }({\mathrm{\&Delta;}}_{\mathrm{\&psi;}}-t^{*})+\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{W}_{\mathrm{ij}}^2{x}_{i\min }({\mathrm{\&Delta;}}_{\mathrm{\&psi;}}+t^{*})}{2{\mathrm{\&Delta;}}_{\mathrm{\&psi;}}}---\left(9\right)$

Will use the time domain center and the radius of female small echo in the following formula, the center that can calculate the time-domain window of Morlet small echo according to the definition of Wavelet time-frequency parameter is 0, and radius is 0.7071.

The initial setting up of

And for the initial setting up of hidden layer to the weights of output layer, can accomplish through following steps:

1. produce the initial setting up value of equally distributed random number as

on [1,1] interval at first at random;

2. then

carried out normalization by row:

${\mathrm{<figure-callout\; id="1"\; label="W\; ki"\; filenames="HSA00000584101000031.png,HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ki}}^{}=\frac{{\mathrm{<figure-callout\; id="1"\; label="W\; ki"\; filenames="HSA00000584101000031.png,HSA00000584101000032.png"\; state="\{\{state\}\}">W</figure-callout>}}_{\mathrm{ki}}^{}}{\sqrt{\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\left(W_{\mathrm{ki}}^1\right)}^2}},k=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,p---\left(10\right)$

Step 106 makes up recurrence small echo Elman network model;

Step 107 initialization weights also carry out the auto-correlation correction;

Step 108 is utilized genetic algorithm training small echo Elman network;

Judge whether small echo Elman network satisfies the setting demand, if, execution in step 109; Otherwise adjustment training parameter.

The generalization ability of step 109 checking recurrence small echo Elman network model;

Judge whether generalization ability reaches requirement, is, execution in step 110; Otherwise, return execution in step 108.

After step 110 reaches requirement, storage recurrence small echo Elman network model parameter.

Above-mentioned braking ability parameter comprises: parameter such as wheel braking retarded velocity, damping force, brake power are poor, wheel slip, car load braking deceleration and car load damping force.

Shown in Figure 2 is based on the hereditary training algorithm flow process in the recurrence small echo Elman network machine motor-car safety monitoring system braking ability discrimination method, may further comprise the steps:

The contraction-expansion factor and the shift factor of step 1081 pair weights, threshold value, wavelet function are carried out real coding;

Step 1082 produces an initial population between [1,1] at random;

The error function of step 1083 computational grid, the functional value of the fitness of definite individuality, error is big more, and adaptive value is more little;

Step 1084 is then changeed step 1087 if end condition satisfies;

Step 1085 selects the maximum individuality of some fitness directly to entail the next generation, and all the other probability of confirming by adaptive value carry out heredity;

Step 1086 is intersected and mutation operation by certain probability, produces population of future generation;

Step 1087 stops circulation, obtains best chromosome.Be converted into the contraction-expansion factor and the shift factor of weights, threshold value, wavelet function then.

Be illustrated in figure 3 as recurrence small echo Elman network structure, the Morlet small echo excitation function that recurrence small echo Elman network structure adopts, each hidden neuron all has two additional parameter a _i, b _i, represent the contraction-expansion factor and the shift factor of small echo action function respectively.This recurrence small echo Elman neural network has n+1 input neuron, and p output neuron accepted a layer neuron for m hidden neuron and m.x ₀(t)=1, the network weight of correspondence

Representing the threshold value of hidden neuron i, is not actual outside input.H ₀(t)=1, map network weights

Representing the threshold value of output neuron k, is not actual outside input.

x _i(t) ∈ R ⁿInscribe recurrence small echo Elman neural network n dimension input vector during for t;

y _k(t) ∈ R ^pInscribe the p dimension output vector of recurrence small echo Elman neural network during for t;

H _i(t) ∈ R ^mBe hidden layer output;

x _c(t) ∈ R ^mFor accepting layer output.

If feedback gain is α, promptly each accept constantly the layer neuron be output as:

x _c，l(t)＝αH _l(t-1) (11)

The dynamic equation of these network input and output this moment can be expressed as:

Wherein:

be hidden neuron i with output layer neuron k between be connected weights;

be input layer j with hidden neuron i between be connected weights;

is for accepting the weights that are connected between layer neuron l and the hidden neuron i;

H _i(t) be the output of hidden neuron i;

Be wavelet function, a _iBe the coefficient of dilatation of small echo, b _iTranslation coefficient for small echo.

Hidden layer neuron number: i=1,2 ..., m; Input layer number: j=0,1,2 ..., n,

Accept layer neuron number a: l=1,2 ..., m; The neuron number of output layer: k=1,2 ..., p.

Shown in Fig. 4 a, 4b and 4c, implement platform in kind based on the motor vehicle safety monitoring system braking ability discrimination method of recurrence small echo Elman network.Fig. 4 a comprises central control module 3 in wheel 1, wheel intelligent sensing module 2, the car; Wherein wheel intelligent sensing module comprises three accelerometers, a pressure transducer and a temperature sensor; And wheel intelligent sensing module is installed on the surface, wheel hub equator of each wheel, and central control module 3 is installed in the car in the car; Realize two-way communication through the Zigbee wireless radio frequency mode between the central control module 3 in wheel intelligent sensing module 2 and the car.Stain is four three-dimensional acceleration intelligent sensings of vehicle body module installation site among Fig. 4 b, adopts square four apex configuration schemes.Fig. 4 c is the one dimension normal acceleration meter that is installed in one of them suspension of motor vehicle.

The above is merely preferred embodiment of the present invention, and is in order to restriction the present invention, not all within spirit of the present invention and principle, any modification of being done, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1. motor vehicle safety monitoring system braking ability discrimination method is characterized in that, said method comprises:

A obtains acceleration, pressure, temperature input and braking ability parameter output function variable;

B eliminates data redundancy through redundant information relation between the PCA analysis operation variable;

C confirms recurrence small echo ELman neural network structure, and obtains a plurality of braking ability of operational process accurately parameter values through discrimination method.

2. motor vehicle safety monitoring system braking ability discrimination method according to claim 1 is characterized in that said step C specifically comprises:

Initialization recurrence small echo ELman neural network model parameter;

Make up recurrence small echo ELman network model;

The initialization weights also carry out the auto-correlation correction;

Utilize genetic algorithm training small echo ELman network;

The generalization ability of checking recurrence small echo ELman network model;

Storage recurrence small echo ELman network model parameter.

3. motor vehicle safety monitoring system braking ability discrimination method according to claim 2 is characterized in that, said recurrence small echo ELman network structure adopts the type combination of compacting.

4. motor vehicle safety monitoring system braking ability discrimination method according to claim 2 is characterized in that, said small echo ELman network genetic algorithm comprises:

Contraction-expansion factor and shift factor to weights, threshold value, wavelet function are carried out real coding;

Produce an initial population between [1,1] at random;

The error function of computational grid, the functional value of the fitness of definite individuality;

If end condition satisfies, then stop circulation, obtain chromosome; Otherwise

Select the big individuality of some fitness directly to entail the next generation, all the other probability of confirming by adaptive value carry out heredity;

Intersect and mutation operation by certain probability, produce population of future generation;

Stop circulation, obtain best chromosome, be converted into the contraction-expansion factor and the shift factor of weights, threshold value, wavelet function then.

5. motor vehicle safety monitoring system braking ability discrimination method according to claim 1; It is characterized in that; Confirm also to comprise before the recurrence small echo ELman neural network structure: confirm the hidden layer neuron number, with the excitation function of Morlet small echo as hidden layer neuron; Confirm also to comprise behind the recurrence small echo ELman neural network structure through orthogonal experimental method and obtain experiment training sample data, these training sample data comprise training sample and desired output sample.

6. motor vehicle safety monitoring system braking ability discrimination method according to claim 5; It is characterized in that; Said hidden layer neuron number is chosen and is adopted experimental formula and experiment to choose the method that combines; Confirm the number range of a hidden layer neuron according to experimental formula

; And set and train step number and training precision accordingly; The hidden layer neuron number is confirmed in preference pattern training from small to large then.

Images