CN114548311B - Hydraulic equipment intelligent control system based on artificial intelligence - Google Patents
Hydraulic equipment intelligent control system based on artificial intelligence Download PDFInfo
- Publication number
- CN114548311B CN114548311B CN202210191927.2A CN202210191927A CN114548311B CN 114548311 B CN114548311 B CN 114548311B CN 202210191927 A CN202210191927 A CN 202210191927A CN 114548311 B CN114548311 B CN 114548311B
- Authority
- CN
- China
- Prior art keywords
- neural network
- confidence
- confidence coefficient
- pressure data
- reliability
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to the technical field of neural networks, in particular to an intelligent control system of hydraulic equipment based on artificial intelligence. The system is an artificial intelligent optimization operating system, comprising: the device comprises a neural network training module, a neural network application module, a credibility acquisition module and an intelligent control module. The neural network training module is used for training a state judgment neural network and a confidence coefficient neural network; the neural network application module is used for acquiring the confidence coefficient of the pressure data; the confidence level obtaining module is used for screening out the number proportion of high confidence levels in the confidence level sequence; simultaneously analyzing the sequence variation trend, and acquiring the data reliability according to the variation trend and the quantity ratio; obtaining a reliability coefficient, and further obtaining the reliability of the state judgment neural network; and the intelligent control module is used for intelligently controlling the hydraulic equipment according to the result of the state judgment neural network and the corresponding credibility. The embodiment of the invention can adopt the neural network model to detect the leakage state of the hydraulic equipment in time and carry out intelligent regulation.
Description
Technical Field
The invention relates to the technical field of neural networks, in particular to an intelligent hydraulic equipment control system based on artificial intelligence.
Background
Artificial Neural Networks (ans), also referred to as Neural Networks (NNs) or Connection models (Connection models), are algorithmic mathematical models that Model animal Neural network behavior characteristics and perform distributed parallel information processing. The network achieves the aim of processing information by adjusting the mutual connection relationship among a large number of nodes in the network depending on the complexity of the system.
At present, a neural network model is adopted in a hydraulic control system to adjust parameters or estimate a system state to form a normal state, but the neural network has larger unexplainable performance, and the reliability of a prediction result of the neural network is an uncertain factor in the process of estimating the system state, so that whether an output result of the neural network is accurate or not cannot be judged, further, the intelligent control of hydraulic equipment is not accurate, and the control process is unreasonable.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide an intelligent control system of hydraulic equipment based on artificial intelligence, and the adopted technical scheme is as follows:
one embodiment of the invention provides an intelligent hydraulic equipment control system based on artificial intelligence, which comprises the following modules:
the neural network training module is used for acquiring pressure data of the hydraulic cylinder at each moment and judging a neural network by utilizing a pressure data sequence of a plurality of preset time periods and a corresponding leakage state training state; calculating the confidence corresponding to each pressure data through the system error corresponding to each pressure data, and respectively training a confidence neural network by using the pressure data and the corresponding confidence under different states;
the neural network application module is used for inputting the acquired pressure data into the trained state judgment neural network, outputting a corresponding state, inputting the pressure data into the trained confidence coefficient neural network in the corresponding state, and outputting a corresponding confidence coefficient;
the confidence coefficient acquisition module is used for acquiring a confidence coefficient sequence of a preset time period before the current moment and screening out the number proportion of confidence coefficients larger than a confidence coefficient threshold value; analyzing the variation trend of the confidence sequence, and acquiring the data reliability according to the variation trend and the quantity ratio; converting the confidence sequence before and after screening into a pressure data range, respectively inputting the pressure data range into a state judgment neural network, obtaining a reliability coefficient according to output results of the previous time and the next time, and obtaining the reliability of the state judgment neural network based on the data reliability and the reliability coefficient;
and the intelligent control module is used for evaluating the stability of the intelligent control system of the hydraulic equipment according to the result of the state judgment neural network and the corresponding credibility so as to intelligently control the hydraulic equipment based on the stability.
Preferably, the neural network training module includes:
the confidence coefficient calculation unit is used for acquiring theoretical displacement of a valve core in the hydraulic cylinder corresponding to each pressure data and actual displacement of the hydraulic cylinder, taking a difference value between the theoretical displacement and the actual displacement as the system error, and acquiring the confidence coefficient according to a ratio of the system error to the theoretical displacement; the confidence level is in a negative correlation with the system error.
Preferably, the neural network training module includes:
and the confidence coefficient neural network training unit is used for acquiring the state of the pressure data sequence through the state judgment neural network, dividing the pressure data sequence into different groups according to different states, and training the confidence coefficient neural network in the group of states by using the pressure data and the corresponding confidence coefficient of each group respectively.
Preferably, the reliability obtaining module includes:
and the quantity proportion acquisition unit is used for performing threshold segmentation on the confidence coefficient sequence to acquire the confidence coefficient threshold, screening out the confidence coefficient quantity which is greater than the confidence coefficient threshold in the confidence coefficient sequence, and taking the ratio of the confidence coefficient quantity to the quantity of all elements of the confidence coefficient sequence as the quantity proportion.
Preferably, the reliability obtaining module includes:
and the variation trend analysis unit is used for establishing a coordinate system by taking the time as an abscissa and taking the confidence coefficient corresponding to the pressure data at each time as an ordinate, acquiring coordinate points of the confidence coefficient sequence in the coordinate system, acquiring principal component direction vectors of the coordinate points, judging the increasing and decreasing variation trend according to the direction of the principal component direction vectors, and acquiring the variation amplitude according to the included angle between the principal component direction vectors and the horizontal direction.
Preferably, the reliability obtaining module includes:
a data reliability obtaining unit, configured to obtain the data reliability according to the change amplitude and the number ratio; when the principal component direction vector is in an increasing trend, the variation amplitude and the data reliability are in positive correlation; when the principal component direction vector is in a decreasing trend, the variation amplitude and the data reliability are in a negative correlation relationship.
Preferably, the reliability obtaining module includes:
the reliability coefficient acquisition unit is used for converting the confidence coefficient sequence into a pressure data range, inputting the pressure data range into the state judgment neural network and outputting a corresponding first state; setting elements which are not greater than the confidence coefficient threshold value in the confidence coefficient sequence to zero to obtain a screening sequence, converting the screening sequence into a pressure data range, inputting the pressure data range into the state judgment neural network, and outputting a corresponding second state; and acquiring the reliability coefficient according to whether the first state and the second state are consistent.
Preferably, the intelligent control module includes:
the stability evaluation unit is used for stabilizing the intelligent control system of the hydraulic equipment when the output result of the state judgment neural network is normal and the reliability is greater than the credible threshold value; otherwise, the intelligent control system of the hydraulic equipment is unstable and needs to be intelligently adjusted.
The embodiment of the invention at least has the following beneficial effects:
and calculating the confidence of the state judgment network according to the output result of the confidence network, further evaluating the reliability of the state judgment network, and evaluating the leakage result of the current hydraulic equipment by using the judgment result and the reliability of the state judgment network so as to regulate the pressure of the hydraulic equipment. The embodiment of the invention can adopt the neural network model to detect the leakage state of the hydraulic equipment in time, can make corresponding early warning response through intelligent control, improves the sensitivity of the system, and can predict and evaluate the state of the system more accurately.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a system block diagram of an artificial intelligence based intelligent control system for a hydraulic device according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a position servo system of the valve-controlled hydraulic cylinder.
Detailed Description
In order to further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description, the structure, the features and the effects of the hydraulic device intelligent control system based on artificial intelligence according to the present invention are provided with the accompanying drawings and the preferred embodiments. In the following description, the different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes a specific scheme of the hydraulic equipment intelligent control system based on artificial intelligence in detail with reference to the accompanying drawings.
Referring to fig. 1, a system block diagram of an artificial intelligence based hydraulic device intelligent control system according to an embodiment of the present invention is shown, where the system includes the following modules:
the system comprises a neural network training module 100, a neural network application module 200, a credibility obtaining module 300 and an intelligent control module 400.
The neural network training module 100 is configured to acquire pressure data of the hydraulic cylinder at each moment, and determine a neural network by using a pressure data sequence of a plurality of preset time periods and a corresponding leakage state training state; and calculating the confidence corresponding to the pressure data according to the system error corresponding to each pressure data, and respectively training a confidence neural network by using the pressure data in different states and the corresponding confidence.
Specifically, the neural network training module 100 includes a pressure data obtaining unit 110, a state determining neural network training unit 120, a confidence calculating unit 130, and a confidence neural network training unit 140.
The pressure data acquisition unit 110 is used to acquire pressure data of the hydraulic cylinder at each time.
Pressure data at each moment inside the hydraulic cylinder is obtained by a pressure sensor installed near the hydraulic pump.
The state judgment neural network training unit 120 is configured to train a state judgment neural network using the pressure data sequences of multiple preset time periods and corresponding leakage states.
In historical data, selecting a plurality of pressure data sequences of preset time periods, simultaneously obtaining corresponding leakage states as labels, marking event properties, using 0 and 1 as label data, wherein 0 represents that leakage does not occur, and 1 represents that leakage occurs, training a classification network through a cross entropy loss function, and obtaining a state judgment neural network.
The confidence coefficient calculation unit 130 is configured to obtain a theoretical displacement of a valve element in the hydraulic cylinder corresponding to each pressure data and an actual displacement of the hydraulic cylinder, use a difference between the theoretical displacement and the actual displacement as a system error, and obtain a confidence coefficient according to a ratio of the system error to the theoretical displacement; the confidence level is in a negative correlation with the system error.
According to the principle of the prior art valve-controlled hydraulic cylinder, as shown in fig. 2, the system sends a command signal
Then the control signal is generated by the controller and the servo amplifier to control the displacement of the valve core of the servo valve
The displacement of the piston rod of the hydraulic cylinder is controlled by the displacement of the valve core of the servo valve
Displacement of piston rod
Obtaining a displacement feedback signal after the detection of the displacement sensor
Then with the command signal
Deviation of comparison
And substituted into the controller to regulate the displacement of the piston rod.
Fig. 2 is from the paper "neural network based valve-controlled hydraulic cylinder system fault diagnosis-smelling middle flying", see page 24 of the paper.
The hydraulic cylinder is under the action of thrust, and the system sends command signals
And displacement signal of displacement sensor
Have errors, so that the command signals are often required
Corresponding theoretical piston rod displacement
Actual displacement from the piston rod
The difference value of (2) is used for adjusting the piston rod, but an error still exists after the adjustment, and the error is taken as a system error
I.e. the difference between the theoretical displacement and the real displacement of the piston rod. Meanwhile, the internal pressure change and the position of the piston rod are in positive correlation, so that the pressure error can be further represented by calculating the displacement error.
The displacement value and pressure value of the piston rod change with the change of leakage amount, and the system error
The larger the displacement data, the less trustworthy the displacement data. Obtaining confidence coefficient through system error and theoretical displacement, wherein the larger the error is, the smaller the confidence coefficient is, and the formula for obtaining the confidence coefficient corresponding to the pressure data is as follows:
wherein, s represents the degree of confidence,
the error of the system is represented by,
indicating the theoretical displacement of the piston rod corresponding to the command data.
The smaller the system error proportion is, the larger the confidence coefficient is, and the relationship between the confidence coefficient and the system error is simulated through an inverse proportion function.
The confidence neural network training unit 140 is configured to obtain the state of the pressure data sequence through the state judgment neural network, divide the pressure data sequence into different groups according to different states, and train the confidence neural network in the group of states by using the pressure data of each group and the corresponding confidence.
Inputting historical data into a trained state judgment neural network, acquiring the state of a pressure data sequence, respectively taking the data with leakage and the data with leakage as a group, respectively training confidence coefficient neural networks, taking the pressure data at each moment as a training set when training the confidence coefficient neural networks, taking the corresponding confidence coefficient as a label to train the confidence coefficient neural networks, and taking a loss function as a cross entropy loss function. And finally obtaining a trained first confidence degree neural network of an undiscovered state and a second confidence degree neural network of a leaked state.
The neural network application module 200 is configured to input the collected pressure data into the trained state judgment neural network, output a corresponding state, input the pressure data into the trained confidence neural network in the corresponding state, and output a corresponding confidence.
Specifically, pressure data are collected in real time, a pressure data sequence of a preset time period before the current time is used as the pressure sequence of the current time, and the pressure data sequence is input into a trained state judgment neural network to obtain an output result. And determining which confidence coefficient neural network is selected to carry out confidence coefficient calculation of data according to the output result of the state judgment neural network. If the output result of the stability neural network is 0, indicating that the system has no leakage event, selecting a first confidence coefficient neural network to calculate the confidence coefficient; and if the output result of the stability neural network is 1, indicating that a leakage event occurs, selecting a second confidence coefficient neural network for performing confidence coefficient calculation.
It should be noted that, since the system error is smaller than the theoretical displacement, and the value of the obtained confidence coefficient is smaller than 1, after the confidence coefficient is obtained, all the obtained confidence coefficients are normalized according to the maximum value of the confidence coefficient, and the value range is adjusted to [0,1].
The confidence level obtaining module 300 is configured to obtain a confidence level sequence of a preset time period before the current time, and screen out a number proportion of confidence levels larger than a confidence level threshold; meanwhile, analyzing the variation trend of the confidence coefficient sequence, and acquiring the data reliability according to the variation trend and the number ratio; the confidence sequence before and after screening is converted into a pressure data range to be respectively input into the state judgment neural network, a reliability coefficient is obtained according to the output results of the previous time and the next time, and the reliability of the state judgment neural network is obtained based on the data reliability and the reliability coefficient.
Specifically, the reliability obtaining module 300 includes a number ratio obtaining unit 310, a variation trend analyzing unit 320, a data reliability obtaining unit 330, a data reliability obtaining unit 340, a reliability coefficient obtaining unit 350, and a reliability evaluating unit 360.
The number proportion obtaining unit 310 is configured to obtain a confidence threshold by performing threshold segmentation on the confidence sequence, screen out the confidence number greater than the confidence threshold in the confidence sequence, and use a ratio of the confidence number to the number of all elements of the confidence sequence as the number proportion.
And performing threshold segmentation on the obtained confidence coefficient sequence by a maximum inter-class variance method to obtain an optimal confidence coefficient threshold, screening out confidence coefficients larger than the confidence coefficient threshold from the confidence coefficient sequence as high confidence coefficients, and obtaining the number proportion r of the number of the high confidence coefficients in the confidence coefficient sequence, wherein the larger the number proportion r is, the more the high confidence coefficients in the confidence coefficient sequence are, the more accurate the judgment result of the state judgment neural network is.
And the variation trend analysis unit 320 is configured to establish a coordinate system by using the time as an abscissa and using the confidence corresponding to the pressure data at each time as an ordinate, acquire coordinate points of the confidence sequence in the coordinate system, acquire principal component direction vectors of the coordinate points, judge an increasing and decreasing variation trend of the principal component direction vectors according to directions of the principal component direction vectors, and acquire a variation amplitude of the principal component direction vectors according to an included angle between the principal component direction vectors and a horizontal direction.
And acquiring coordinates of the coordinate points, acquiring principal component directions of the data by using a PCA algorithm, and acquiring K principal component directions, wherein each principal component direction is a 2-dimensional unit vector and corresponds to a characteristic value. The principal direction with the largest eigenvalue is taken as the principal direction of the data, which represents the direction with the largest projection variance of the data, i.e. the principal distribution direction of the data, the principal direction of the data is a vector, and the angle between the vector and the horizontal direction is recorded as
。
Judging the change trend of the sequence according to the angle interval distribution of the included angle between the principal component direction vector and the horizontal direction, if the included angle between the principal component direction vector and the horizontal direction
Falling between 0 deg. -90 deg., the confidence level is increasing. If the principal component direction vector is parallel to the horizontal directionCorner
Falling between-90 deg. -0 deg., the confidence is a decreasing trend.
By the angle between the principal component direction vector and the horizontal direction
The ratio of the absolute value of (a) to 90 DEG is taken as the amplitude of change
：
。
A data reliability obtaining unit 330, configured to obtain data reliability according to the variation amplitude and the number ratio; when the direction vector of the principal component is in an increasing trend, the variation amplitude and the data reliability are in positive correlation; when the principal component direction vector is in a decreasing trend, the variation amplitude and the data reliability are in a negative correlation relationship.
Forming a binary group by the variation trend and the variation amplitude of the principal component direction vector, wherein the first element represents the increasing or decreasing trend, 0 represents the decreasing trend, and 1 represents the increasing trend; the second element being amplitude of variation
I.e. the magnitude of the trend. The representation of the doublet is (0, t) or (1, t).
For example: (1, 0.9), indicating an increasing tendency and the increasing amplitude is 0.9. The closer the magnitude of the change is to 1, the faster the rate of change of the confidence of the data is illustrated.
And a data reliability obtaining unit 340 for obtaining data reliability according to the variation trend and the number ratio.
When the variation trend is increased, 1+ tis taken as the increasing proportion of the number ratio r, and the data reliability of the data sequence
(ii) a When the variation trend is reduced, 1-t is taken as the reduction ratio of the number proportion r, and the data reliability of the data sequence
。
The confidence coefficient under the ideal condition is 1, on the premise of increasing the trend, the larger the increasing trend is, the more quickly the confidence coefficient can be increased to the ideal trend, and the greater the data reliability is; on the premise of reducing the trend, the larger the reducing trend is, the farther the confidence coefficient is from the ideal trend, and the data reliability is lower.
A reliability coefficient obtaining unit 350, configured to convert the confidence sequence into a pressure data range input state judgment neural network, and output a corresponding first state; setting elements not greater than the confidence coefficient threshold value in the confidence coefficient sequence to zero to obtain a screening sequence, converting the screening sequence into a pressure data range input state judgment neural network, and outputting a corresponding second state; and acquiring the reliability coefficient according to whether the first state and the second state are consistent.
And converting the confidence coefficient sequence into a pressure data range in an equal proportion according to the value of the pressure data, using the pressure data range as the input of the state judgment neural network, outputting a first state, wherein if the output state is a leakage-free state, the first state is 0, and if the output state is a leakage state, the first state is 1.
And converting the screening sequences into a pressure data range in the same equal proportion, using the pressure data range as the input of the state judgment neural network, outputting a second state, wherein if the output state is a leakage-free state, the second state is 0, and if the output state is a leakage state, the second state is 1.
If the first state and the second state are the same, the data with small confidence coefficient has small influence on the result reliability, namely the output result of the state judgment neural network is accurate, and the reliability coefficient v is 1; otherwise, the accuracy of the output result of the state judgment neural network is low, and the reliability coefficient v at the moment is 0.8.
The reliability evaluation unit 360 is configured to determine the reliability of the neural network based on the data reliability and the reliability coefficient acquisition state.
And judging the reliability of the neural network by taking the product of the data reliability and the reliability coefficient as a state. The higher the reliability is, the more accurate the judgment result of the state judgment neural network is.
And the intelligent control module 400 is used for evaluating the stability of the intelligent control system of the hydraulic equipment according to the result of the state judgment neural network and the corresponding credibility so as to intelligently control the hydraulic equipment based on the stability.
Specifically, the intelligent control module 400 includes a stability evaluation unit 410.
The stability evaluation unit 410 is configured to, when the state determines that the output result of the neural network is normal and the reliability is greater than the confidence threshold, stabilize the intelligent control system of the hydraulic device; otherwise, the intelligent control system of the hydraulic equipment is unstable and needs to be intelligently adjusted.
When the output result of the state judgment neural network is normal, namely the state is not leaked, the reliability of the state judgment neural network is used as a correction parameter of the result, and if the reliability of the state judgment neural network is greater than a reliability threshold value, the intelligent control system of the hydraulic equipment is stable and operates normally; if the reliability of the state judgment neural network is not greater than the reliability threshold, it is indicated that although the output result of the state judgment neural network is normal, the reliability of the result is not high enough, leakage may be slight at the moment, and the system is not identified, so that timely adjustment is needed, the internal pressure of the hydraulic cylinder is gradually reduced to stop the operation of the hydraulic equipment, and serious leakage is avoided.
When the output result of the state judgment neural network is abnormal, namely a leakage state, if the reliability of the state judgment neural network is greater than a credible threshold, the output result of the neural network is accurate, hydraulic pressure leakage exists at the moment, intelligent control needs to be carried out on hydraulic equipment, the internal pressure of a hydraulic cylinder is gradually reduced to stop the hydraulic equipment from running, and meanwhile, early warning needing manual maintenance is carried out; if the reliability of the state judgment neural network is not greater than the credible threshold value, the hydraulic leakage is indicated, parts of a hydraulic system are possibly damaged, and the operation needs to be stopped immediately and maintained.
As an example, in the embodiment of the present invention, the value of the confidence threshold is 0.8.
In summary, the embodiment of the present invention includes a neural network training module 100, a neural network application module 200, a reliability obtaining module 300, and an intelligent control module 400.
Specifically, the neural network training module is used for acquiring pressure data of the hydraulic cylinder at each moment, and judging the neural network by using pressure data sequences of a plurality of preset time periods and corresponding leakage state training states; calculating the confidence corresponding to the pressure data according to the system error corresponding to each pressure data, and respectively training a confidence neural network by using the pressure data in different states and the corresponding confidence thereof; the neural network application module is used for inputting the acquired pressure data into the trained state judgment neural network, outputting a corresponding state, inputting the pressure data into the trained confidence coefficient neural network of the corresponding state, and outputting a corresponding confidence coefficient; the confidence coefficient acquisition module is used for acquiring a confidence coefficient sequence of a preset time period before the current time, and screening out the number proportion of confidence coefficients larger than a confidence coefficient threshold value; meanwhile, analyzing the variation trend of the confidence coefficient sequence, and acquiring the data reliability according to the variation trend and the number ratio; converting confidence sequence before and after screening into pressure data ranges, respectively inputting the pressure data ranges into a state judgment neural network, obtaining a reliability coefficient according to output results of the previous time and the next time, and obtaining the reliability of the state judgment neural network based on data reliability and the reliability coefficient; the intelligent control module is used for evaluating the stability of the intelligent control system of the hydraulic equipment according to the result of the state judgment neural network and the corresponding credibility so as to intelligently control the hydraulic equipment based on the stability. The embodiment of the invention can adopt the neural network model to detect the leakage state of the hydraulic equipment in time, can make corresponding early warning response through intelligent control, improves the sensitivity of the system, and can predict and evaluate the state of the system more accurately.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and should not be taken as limiting the scope of the present invention, which is intended to cover any modifications, equivalents, improvements, etc. within the spirit and scope of the present invention.
Claims (6)
1. Hydraulic equipment intelligence control system based on artificial intelligence, its characterized in that, this system includes the following module:
the neural network training module is used for acquiring pressure data of the hydraulic cylinder at each moment and judging a neural network by utilizing a plurality of pressure data sequences of preset time periods and corresponding leakage state training states; calculating the confidence corresponding to each pressure data through the system error corresponding to each pressure data, and respectively training a confidence neural network by using the pressure data and the corresponding confidence under different states;
the neural network application module is used for inputting the acquired pressure data into the trained state judgment neural network, outputting a corresponding state, inputting the pressure data into the trained confidence coefficient neural network in the corresponding state, and outputting a corresponding confidence coefficient;
the confidence coefficient acquisition module is used for acquiring a confidence coefficient sequence of a preset time period before the current moment and screening out the number proportion of confidence coefficients larger than a confidence coefficient threshold value; simultaneously analyzing the variation trend of the confidence coefficient sequence, and acquiring the data reliability according to the variation trend and the quantity ratio; converting the confidence sequence before and after screening into a pressure data range, respectively inputting the pressure data range into a state judgment neural network, obtaining a reliability coefficient according to the output results of the previous time and the next time, and obtaining the reliability of the state judgment neural network based on the data reliability and the reliability coefficient;
the intelligent control module is used for evaluating the stability of the intelligent control system of the hydraulic equipment according to the result of the state judgment neural network and the corresponding credibility so as to intelligently control the hydraulic equipment based on the stability;
the credibility obtaining module comprises:
the variation trend analysis unit is used for establishing a coordinate system by taking the time as an abscissa and taking the confidence coefficient corresponding to the pressure data at each time as an ordinate, acquiring coordinate points of the confidence coefficient sequence in the coordinate system, acquiring principal component direction vectors of the coordinate points, judging the increasing and decreasing variation trend of the principal component direction vectors according to the directions of the principal component direction vectors, and acquiring the variation amplitude of the principal component direction vectors according to the included angle between the principal component direction vectors and the horizontal direction;
a data reliability obtaining unit, configured to obtain the data reliability according to the change amplitude and the number ratio; when the principal component direction vector is in an increasing trend, the variation amplitude and the data reliability are in positive correlation relationship, and the data reliability is
(ii) a When the principal component direction vector is in a decreasing trend, the variation amplitude and the data reliability are in a negative correlation relationship, and the data reliability is
(ii) a Wherein t represents the variation amplitude and r represents the number fraction.
2. The artificial intelligence based hydraulic device intelligence control system of claim 1, wherein the neural network training module comprises:
the confidence coefficient calculation unit is used for acquiring theoretical displacement of a valve core in the hydraulic cylinder corresponding to each pressure data and actual displacement of the hydraulic cylinder, taking a difference value between the theoretical displacement and the actual displacement as the system error, and acquiring the confidence coefficient according to a ratio of the system error to the theoretical displacement; the confidence level is in a negative correlation with the system error.
3. The artificial intelligence based hydraulic device intelligence control system of claim 1, wherein the neural network training module comprises:
and the confidence coefficient neural network training unit is used for acquiring the state of the pressure data sequence through the state judgment neural network, distinguishing the pressure data sequence into different groups according to different states, and training the confidence coefficient neural network in the group of states by using the pressure data and the corresponding confidence coefficient of each group respectively.
4. The artificial intelligence based hydraulic device intelligence control system of claim 1, wherein the confidence level acquisition module includes:
and the quantity proportion acquisition unit is used for performing threshold segmentation on the confidence coefficient sequence to acquire the confidence coefficient threshold, screening out the confidence coefficient quantity which is greater than the confidence coefficient threshold in the confidence coefficient sequence, and taking the ratio of the confidence coefficient quantity to the quantity of all elements of the confidence coefficient sequence as the quantity proportion.
5. The artificial intelligence based hydraulic device intelligence control system of claim 1, wherein the confidence level acquisition module comprises:
the reliability coefficient acquisition unit is used for converting the confidence coefficient sequence into a pressure data range, inputting the pressure data range into the state judgment neural network and outputting a corresponding first state; setting elements which are not greater than the confidence coefficient threshold value in the confidence coefficient sequence to zero to obtain a screening sequence, converting the screening sequence into a pressure data range, inputting the pressure data range into the state judgment neural network, and outputting a corresponding second state; and acquiring the reliability coefficient according to whether the first state and the second state are consistent.
6. The artificial intelligence based hydraulic device intelligence control system of claim 1, wherein the intelligence control module comprises:
the stability evaluation unit is used for stabilizing the intelligent control system of the hydraulic equipment when the output result of the state judgment neural network is normal and the reliability is greater than the credible threshold value; otherwise, the intelligent control system of the hydraulic equipment is unstable and needs to be intelligently adjusted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210191927.2A CN114548311B (en) | 2022-02-28 | 2022-02-28 | Hydraulic equipment intelligent control system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210191927.2A CN114548311B (en) | 2022-02-28 | 2022-02-28 | Hydraulic equipment intelligent control system based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114548311A CN114548311A (en) | 2022-05-27 |
| CN114548311B true CN114548311B (en) | 2022-12-02 |
Family
ID=81660790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210191927.2A Active CN114548311B (en) | 2022-02-28 | 2022-02-28 | Hydraulic equipment intelligent control system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114548311B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116503229B (en) * | 2023-06-27 | 2023-09-26 | 成都秦川物联网科技股份有限公司 | Intelligent gas pipe network inspection method, internet of things system and storage medium |
| CN116717734B (en) * | 2023-08-10 | 2023-10-20 | 牧马人(山东)勘察测绘集团有限公司 | Data monitoring method for steady-state operation of natural gas pipeline |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010055303A (en) * | 2008-08-27 | 2010-03-11 | Denso It Laboratory Inc | Learning data management device, learning data management method and air-conditioner for vehicle, and control device of apparatus |
| CN110874685A (en) * | 2018-09-04 | 2020-03-10 | 中国计量科学研究院 | Intelligent electric energy meter running state distinguishing method and system based on neural network |
| CN113872328A (en) * | 2021-09-26 | 2021-12-31 | 国网安徽省电力有限公司铜陵供电公司 | Transformer substation remote intelligent inspection method and system based on neural network |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3493154A1 (en) * | 2017-12-01 | 2019-06-05 | Koninklijke Philips N.V. | Segmentation system for segmenting an object in an image |
| LU100902B1 (en) * | 2018-08-09 | 2020-02-17 | Iee Sa | System including a device and a learning algorithm to perform a confidence measure-based classification or regression problem |
| CN109800537B (en) * | 2019-02-20 | 2022-11-18 | 大连理工大学 | Machine tool thermal error model reliability calculation method based on deep neural network and Monte Carlo method |
| CN109919302B (en) * | 2019-02-28 | 2021-11-02 | 新华三技术有限公司 | Training method and device for neural network of image |
| CN110848896A (en) * | 2019-11-12 | 2020-02-28 | 深圳孚沃德斯科技有限公司 | Intelligent energy-saving control system and method for air conditioner cooling system based on neural network |
| US20220004863A1 (en) * | 2020-07-01 | 2022-01-06 | International Business Machines Corporation | Confidence classifiers for diagnostic training data |
| CN112991418B (en) * | 2021-03-09 | 2024-03-29 | 北京地平线信息技术有限公司 | Image depth prediction and neural network training method and device, medium and equipment |
-
2022
- 2022-02-28 CN CN202210191927.2A patent/CN114548311B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010055303A (en) * | 2008-08-27 | 2010-03-11 | Denso It Laboratory Inc | Learning data management device, learning data management method and air-conditioner for vehicle, and control device of apparatus |
| CN110874685A (en) * | 2018-09-04 | 2020-03-10 | 中国计量科学研究院 | Intelligent electric energy meter running state distinguishing method and system based on neural network |
| CN113872328A (en) * | 2021-09-26 | 2021-12-31 | 国网安徽省电力有限公司铜陵供电公司 | Transformer substation remote intelligent inspection method and system based on neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114548311A (en) | 2022-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114548311B (en) | Hydraulic equipment intelligent control system based on artificial intelligence | |
| CN108898215B (en) | Intelligent sludge bulking identification method based on two-type fuzzy neural network | |
| CN111444953B (en) | Sensor fault monitoring method based on improved particle swarm optimization algorithm | |
| CN112966891A (en) | River water environment quality prediction method | |
| CN115016276B (en) | Intelligent water content adjustment and environment parameter Internet of things big data system | |
| CN115654208B (en) | Adjusting valve opening monitoring method and system based on image recognition | |
| WO2021114320A1 (en) | Wastewater treatment process fault monitoring method using oica-rnn fusion model | |
| CN117495205B (en) | Industrial Internet experiment system and method | |
| CN112651444B (en) | Self-learning-based non-stationary process anomaly detection method | |
| CN116865343B (en) | Model-free self-adaptive control method, device and medium for distributed photovoltaic power distribution network | |
| CN116664846B (en) | Method and system for realizing 3D printing bridge deck construction quality monitoring based on semantic segmentation | |
| CN110889646B (en) | Method for checking correctness of operating parameters of generator set adapting to working condition change | |
| CN112685857A (en) | Bearing residual life prediction method based on transfer migration | |
| CN117094608A (en) | Titanium belt production control method and system combining application requirements | |
| CN109725526B (en) | Multivariable semi-adaptive prediction control method | |
| CN110045716B (en) | Method and system for detecting and diagnosing early fault of closed-loop control system | |
| CN116055516A (en) | Automatic monitoring system based on industrial Internet of things | |
| CN109697330A (en) | A kind of fault detection system design for information physical system | |
| CN114880930A (en) | Entity equipment digital twin modeling simulation prediction method based on online learning | |
| CN114415503A (en) | Temperature big data thing networking detects and intelligent control system | |
| CN103513648B (en) | The performance test methods of control system and system | |
| KR20010036423A (en) | Method and apparatus for detecting fault using General Regression Neural Network | |
| CN110244563A (en) | A kind of identification of neural Networks Internal Model Control device model mismatch and online updating method | |
| CN110676855A (en) | Intelligent optimization and adjustment method for reactive voltage control parameters of power distribution network | |
| CN108764337A (en) | Underwater robot propeller fault degree discrimination method based on relative grey correlative degree boundary constraint |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |