CN105045091A - Dredging process intelligent decision analysis method based on fuzzy neural control system - Google Patents

Abstract

The invention discloses a dredging process intelligent decision analysis method based on a fuzzy neural control system. The method comprises the steps that 1, data influencing related decision parameters of a dredging construction technology are collected; 2, a matlab principal component analysis method is used to find out the feature thresholds of several decision parameters with the maximum contribution rate; 3, a knowledge base is established; 4, a fuzzy neural control system mechanism is selected and artificial neurons are established, and the number of hidden layers of a neural network is determined according to the dredging complexity; and 5, the fuzzy neural control system is established to carry out assisted decision on dredging construction. According to the invention, the space complexity of the neural network and fuzzy control time complexity are fused, so that the degree of dredging automation is high.

Description

Based on the dredging technique intelligent decision analytical approach of Fuzzy Neural Control system

Technical field

The present invention relates to a kind of dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system, belong to dredging work field.

Background technology

Dredging is as underwater operation, and process control parameter is numerous, and current domestic dredging automaticity is not high, still based on manual operation.Even if veteran operating personnel, because the factor affecting mud flow state is intricate, the settling velocity of such as concentration of hydraulic mixture, mud speed rate, sediment grain size, different silt and tube performance etc., and influence each other between each factor and cause qualitative test relative difficulty.Cause dredging to be produced and be in low yield poor efficiency and high energy consumption maximum discharge state always.Therefore, improve dredging automaticity and seem particularly urgent in today.

Nineteen sixty-five, U.S. autonetics person Z.A.Zadeh proposes fuzzy set concept, and pioneering fuzzy set theory, for describing the phenomenon not having clear and definite boundary and fuzzy extension.Nineteen forty-three, French psychologist W.S.McCuloch and W.Pitts proposes neuron models.Merge the time complexity of fuzzy control by the space complexity of neural network, both combinations have just been born Fuzzy Neural Control system, and Fuzzy Neural Control system is we provide a kind of new thinking.

Summary of the invention

Present invention achieves a kind of dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system, merged the time complexity of fuzzy control by the space complexity of neural network, make dredging operation automaticity higher.

In order to achieve the above object, the technical solution adopted in the present invention is:

Based on the dredging technique intelligent decision analytical approach of Fuzzy Neural Control system, comprise the following steps,

Step one, collects the data information affecting dredging operation construction technology relevant Decision parameter;

Step 2, application matlab principal component analytical method, finds out several its characteristic threshold value of decision-making parameter that contribution rate is maximum;

Step 3, sets up knowledge base; Described knowledge base comprises fuzzy rule base and fuzzy database;

Fuzzy rule is defined in described fuzzy rule base; The subordinate function used in fuzzy rule is defined in described fuzzy database;

Step 4, selects Fuzzy Neural Control system authority, sets up artificial neuron, according to the hidden layer number of dredging operation complexity determination neural network;

Step 5, sets up Fuzzy Neural Control system, carries out aid decision making to dredging operation construction.

Affect that dredging operation construction technology relevant Decision parameter comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, flow rate of pipelines, reamer transverse moving speed, reamer cut mud thickness, reamer forward travel distance, reamer depth, chassis stroke, the equal concentration of pipeline and exit velocity.

The maximum decision-making parameter of contribution rate comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, reamer transverse moving speed and reamer forward travel distance.

Described subordinate function adopts threefold division to calculate and obtains, and detailed process is,

1) definition space Ω, is divided into three sub spaces A by Ω ₁, A ₂and A ₃;

2) A is defined ₁and A ₂separation/face be ξ, definition A ₂and A ₃separation/face be η;

3) use the formula of the subordinate function of split plot design calculating immediately as follows,

If (ξ, η) is the one group of continuous vector immediately meeting P (ξ, η)=1, if corresponding one of each value of (ξ, η) maps e, have

e(ξ，η):Ω→U＝{A ₁,A ₂,A ₃}

Namely three corresponding subspaces can all be drawn after often determining a border

And

$e(\mathrm{\&xi;},\mathrm{\&eta;})\left(x\right)=\left\{\begin{array}{cc}{A}_1& x\&le;\mathrm{\&xi;}\\ A_2& \mathrm{\&xi;}<x\&le;\mathrm{\&eta;}\\ A_3& x>\mathrm{\&eta;}\end{array}\right.$

Wherein: U comprises three subset A ₁, A ₂, A ₃complete or collected works; E (ξ, η) represents that corresponding one of each value of (ξ, η) maps e; E (ξ, η) (x) represents the A when ξ and η determines ₁, A ₂, A ₃span;

Three subordinate functions that then three fuzzy subsets are corresponding with for,

${\mathrm{\&mu;}}_{A_1}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&xi;}}\left(u\right)du$

${\mathrm{\&mu;}}_{A_2}\left(x\right)=1-{\mathrm{\&mu;}}_{A1}\left(x\right)-{\mathrm{\&mu;}}_{A3}\left(x\right)$

${\mathrm{\&mu;}}_{A_3}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&eta;}}\left(u\right)du$

Wherein, P _ξ(u) and P _ηu () is respectively the marginal distribution density function of ξ and η.

Adopt BP algorithm, set up artificial neuron, according to the hidden layer number of dredging operation complexity determination neural network.

Using fuzzy rule base as hidden node, set up Fuzzy Neural Control system, by continuous learning and training until output error is reduced to the degree of permission.

The beneficial effect that the present invention reaches: the present invention merges the time complexity of fuzzy control by the space complexity of neural network, and Fuzzy Neural Control has stronger adaptivity and robustness, simulation result shows, quiet, dynamic property and interference free performance are all better than Traditional PID and conventional fuzzy control, and this can make dredging operation automaticity higher.

Accompanying drawing explanation

Fig. 1 is process flow diagram of the present invention.

Fig. 2 is Fuzzy Neural Control system basic block diagram.

Embodiment

Below in conjunction with accompanying drawing, the invention will be further described.Following examples only for technical scheme of the present invention is clearly described, and can not limit the scope of the invention with this.

As shown in Figure 1, based on the dredging technique intelligent decision analytical approach of Fuzzy Neural Control system, comprise the following steps:

Step one, collects the data information affecting dredging operation construction technology relevant Decision parameter.

Affect dredging operation construction technology relevant Decision parameter, need obtain according to construction experience, the decision-making parameter here collected comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, flow rate of pipelines, reamer transverse moving speed, reamer cut mud thickness, reamer forward travel distance, reamer depth, chassis stroke, the equal concentration of pipeline and exit velocity.

Step 2, application matlab principal component analytical method, finds out several its characteristic threshold value of decision-making parameter that contribution rate is maximum.

Application matlab principal component analytical method, study each decision-making parameter to the contribution rate of output and energy consumption and contribution rate of accumulative total, draw the major component and characteristic threshold value thereof that affect output and energy consumption, theoretical foundation is provided for reducing regulation and control parameter, for the parameter regulation and control of dredging process decision provide the codomain of reasonable, reach the complicacy reducing dredging process decision.

Here affect the major component of output and energy consumption, the decision-making parameter that namely contribution rate is maximum comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, reamer transverse moving speed and reamer forward travel distance.

Step 3, sets up knowledge base.

Knowledge base comprises fuzzy rule base and fuzzy database; Fuzzy rule is defined in fuzzy rule base; The subordinate function used in fuzzy rule is defined in fuzzy database.

Subordinate function adopts threefold division to calculate and obtains, and detailed process is,

A1) definition space Ω, is divided into three sub spaces A by Ω ₁, A ₂and A ₃;

A2) A is defined ₁and A ₂separation/face be ξ, definition A ₂and A ₃separation/face be η;

3) use the formula of the subordinate function of split plot design calculating immediately as follows,

If (ξ, η) is the one group of continuous vector immediately meeting P (ξ, η)=1, if corresponding one of each value of (ξ, η) maps e; Have

e(ξ，η):Ω→U＝{A ₁,A ₂,A ₃}

Namely three corresponding subspaces can all be drawn after often determining a border.

And

$e(\mathrm{\&xi;},\mathrm{\&eta;})\left(x\right)=\left\{\begin{array}{cc}{A}_1& x\&le;\mathrm{\&xi;}\\ A_2& \mathrm{\&xi;}<x\&le;\mathrm{\&eta;}\\ A_3& x>\mathrm{\&eta;}\end{array}\right.$

Wherein: U comprises three subset A ₁, A ₂, A ₃complete or collected works; E (ξ, η) represents that corresponding one of each value of (ξ, η) maps e; E (ξ, η) (x) represents the A when ξ and η determines ₁, A ₂, A ₃span;

Three subordinate functions that then three fuzzy subsets are corresponding with for,

${\mathrm{\&mu;}}_{A_1}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&xi;}}\left(u\right)du$

${\mathrm{\&mu;}}_{A_2}\left(x\right)=1-{\mathrm{\&mu;}}_{A1}\left(x\right)-{\mathrm{\&mu;}}_{A3}\left(x\right)$

${\mathrm{\&mu;}}_{A_3}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&eta;}}\left(u\right)du$

Wherein, P _ξ(u) and P _ηu () is respectively the marginal distribution density function of ξ and η.

Therefore several input main parameters is divided into A ₁, A ₂, A ₃three Estate:

Concentration of hydraulic mixture: less than normal, medium, bigger than normal;

Reamer rotating speed: partially slow, medium, fast;

Dredge pump rotating speed: partially slow, medium, fast;

Reamer transverse moving speed: partially slow, medium, fast;

Reamer forward travel distance: less than normal, medium, bigger than normal;

Such as concentration of hydraulic mixture, during beginning, when exporting concentration value and being less than input value, input setting concentration value, thinks that concentration of hydraulic mixture is less than normal, and be then for medium when output valve equals input value just, when output valve is greater than input value then for bigger than normal, other parameters in like manner.

Concrete subordinate function can be obtained by required parameter value.

Step 4, selects Fuzzy Neural Control system authority, adopts BP algorithm, sets up artificial neuron, according to the hidden layer number of dredging operation complexity determination neural network.

Step 5, sets up Fuzzy Neural Control system, carries out aid decision making to dredging operation construction.

This Fuzzy Neural Control system, using fuzzy rule base as hidden node, sets up Fuzzy Neural Control system, by continuous learning and training until output error is reduced to the degree of permission.

Set up Fuzzy Neural Control system as shown in Figure 2, wherein x ₁, x ₂, x ₃x _nfor input parameter, y ₁, y ₂, y ₃y _nfor output parameter, hidden layer number is one deck, the corresponding fuzzy rule base of hidden node.

If n=5, i.e. x ₁, x ₂, x ₃x ₅be respectively concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, reamer transverse moving speed and reamer forward travel distance, y ₁, y ₂, y ₃y ₅for the actual parameter exported, concentration value that high yield is can be set and the value of other each parameters under this concentration value before carrying out dredging operation, when concentration of hydraulic mixture is higher or lower than setting concentration value, other parameter also can regulate its output quantity automatically, maintains concentration of hydraulic mixture and to reach or close to setting value.

Above-mentioned, based on the dredging technique intelligent decision analytical approach of Fuzzy Neural Control system, merged the time complexity of fuzzy control by the space complexity of neural network, make dredging operation automaticity higher.

The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the prerequisite not departing from the technology of the present invention principle; can also make some improvement and distortion, these improve and distortion also should be considered as protection scope of the present invention.

Claims (6)

1., based on the dredging technique intelligent decision analytical approach of Fuzzy Neural Control system, it is characterized in that: comprise the following steps,

Step one, collects the data information affecting dredging operation construction technology relevant Decision parameter;

Step 2, application matlab principal component analytical method, finds out several its characteristic threshold value of decision-making parameter that contribution rate is maximum;

Step 3, sets up knowledge base; Described knowledge base comprises fuzzy rule base and fuzzy database;

Fuzzy rule is defined in described fuzzy rule base; The subordinate function used in fuzzy rule is defined in described fuzzy database;

Step 4, selects Fuzzy Neural Control system authority, sets up artificial neuron, according to the hidden layer number of dredging operation complexity determination neural network;

Step 5, sets up Fuzzy Neural Control system, carries out aid decision making to dredging operation construction.

2. the dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system according to claim 1, is characterized in that: affect that dredging operation construction technology relevant Decision parameter comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, flow rate of pipelines, reamer transverse moving speed, reamer cut mud thickness, reamer forward travel distance, reamer depth, chassis stroke, the equal concentration of pipeline and exit velocity.

3. the dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system according to claim 2, is characterized in that: the maximum decision-making parameter of contribution rate comprises concentration of hydraulic mixture, reamer rotating speed, dredge pump rotating speed, reamer transverse moving speed and reamer forward travel distance.

4. the dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system according to claim 1, is characterized in that: described subordinate function adopts threefold division to calculate and obtains, and detailed process is,

1) definition space Ω, is divided into three sub spaces A by Ω ₁, A ₂and A ₃;

2) A is defined ₁and A ₂separation/face be ξ, definition A ₂and A ₃separation/face be η;

3) use the formula of the subordinate function of split plot design calculating immediately as follows,

If (ξ, η) is the one group of continuous random vector meeting P (ξ, η)=1, suppose that again the corresponding mapping e of each value of (ξ, η) has

e(ξ，η):Ω→U＝{A ₁,A ₂,A ₃}

Namely three corresponding subspaces can all be drawn after often determining a border

And

$e(\mathrm{\&xi;},\mathrm{\&eta;})\left(x\right)=\left\{\begin{array}{cc}{A}_1& x\&le;\mathrm{\&xi;}\\ A_2& \mathrm{\&xi;}<x\&le;\mathrm{\&eta;}\\ A_3& x>\mathrm{\&eta;}\end{array}\right.$

Wherein: U comprises three subset A ₁, A ₂, A ₃complete or collected works; E (ξ, η) represents that corresponding one of each value of (ξ, η) maps e; E (ξ, η) (x) represents the A when ξ and η determines ₁, A ₂, A ₃span;

Three subordinate functions that then three fuzzy subsets are corresponding with for,

${\mathrm{\&mu;}}_{A_1}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&xi;}}\left(u\right)du$

${\mathrm{\&mu;}}_{A_2}\left(x\right)=1-{\mathrm{\&mu;}}_{A1}\left(x\right)-{\mathrm{\&mu;}}_{A3}\left(x\right)$

${\mathrm{\&mu;}}_{A_3}\left(x\right)={\&Integral;}_x^{\mathrm{\&infin;}}{P}_{\mathrm{\&eta;}}\left(u\right)du$

Wherein, P _ξ(u) and P _ηu () is respectively the marginal distribution density function of ξ and η.

5. the dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system according to claim 1, is characterized in that: adopt BP algorithm, set up artificial neuron, according to the hidden layer number of dredging operation complexity determination neural network.

6. the dredging technique intelligent decision analytical approach based on Fuzzy Neural Control system according to claim 1, it is characterized in that: using fuzzy rule base as hidden node, set up Fuzzy Neural Control system, by continuous learning and training until output error is reduced to the degree of permission.