Hydrostatic rotary table intelligent monitoring system and method

Abstract

The application discloses a kind of static pressure rotary table intelligent monitoring systems and methods, including data acquisition system, big data computing system, control system and artificial intelligence system. Through data acquisition system, rotary table parameter is transferred to big data computing system, control system and artificial intelligence system, and the working state of static pressure rotary table is monitored and evaluated in real time dynamically. Static pressure rotary table and circular oil pad model are established, and each sensing unit is controlled by control system to record the real-time parameters of rotary table;Through data acquisition system, the relevant parameters affecting static pressure rotary table are collected, and the collected data are transmitted into big data computing system for data classification processing;MATLAB is used to build artificial intelligence system, the processed data are input into artificial neural network model for training, and the working condition of static pressure rotary table is evaluated and monitored through output value;The application can monitor the working condition of static pressure rotary table in real time dynamically, and more comprehensively and accurately understand the working performance and condition of static pressure rotary table.

Description

TECHNICAL FIELD

[0001] The application relates to an intelligent monitoring system and method for a hydrostatic rotary table, in particular to an intelligent monitoring system and method for a hydrostatic rotary table based on an artificial neural network, and belongs to the technical field of intelligent monitoring of hydrostatic rotary tables. BACKGROUND

[0002] In an ultra-precision machining device, a hydrostatic rotary table uses a pressure oil film as a medium and injects the pressure oil film into an oil cavity. When the rotary table rotates, the liquid static pressure generated by the flow of the oil film can support the rotary table and make it rotate, so that the rotary table can drive a workpiece to realize the function of ultra-precision machining. The hydrostatic rotary table has been widely applied in heavy machinery and ultra-precision machining due to its large bearing capacity, high rigidity, sensitive operation and high rotation accuracy.

[0003] The hydrostatic rotary table is affected by many complex factors during work. The traditional manual management cannot fully meet the monitoring requirements, and the current monitoring work has the defects of high cost and long cycle. Therefore, an effective monitoring system and method are needed to intelligently monitor and evaluate the working performance of the hydrostatic rotary table. SUMMARY

[0004] The application aims to provide an intelligent monitoring system and method for a hydrostatic rotary table, which can monitor and intelligently control the working performance of the hydrostatic rotary table, improve the intelligentization, automation and maintenance level, monitor the safety hazards through an artificial neural network, and further improve the working efficiency of the hydrostatic rotary table and avoid the occurrence of safety hazards.

[0005] To achieve the above-mentioned target, the technical scheme adopted by the application is an intelligent monitoring system for a hydrostatic rotary table, which comprises a hydraulic pump, a data acquisition system and a big data calculation system. The hydraulic pump is connected with the big data calculation system through the data acquisition system. The data acquisition system mainly comprises an oil tank, a flow sensor, a pressure sensor, a displacement sensor, a temperature sensor, a circular oil pad, a rotary table and an oil pipe. The rotary table is a rotating mechanism of the hydrostatic rotary table, and the bottom thereof is supported by the circular oil pad. The circular oil pad is provided with the displacement sensor and the temperature sensor. The circular oil pad is supplied with oil by the oil tank, and the circular oil pad is connected with the oil tank through the oil pipe. The oil pipe is provided with the flow sensor and the pressure sensor. The working state of the hydrostatic rotary table is monitored and evaluated in real time and dynamically, and the working efficiency and the intelligent management level are further improved.

[0006] The application further provides an intelligent monitoring method for a hydrostatic rotary table, and the implementation process is as follows:

[0007] (1) a model of the hydrostatic rotary table and the circular oil pad is established, and each sensing unit records real-time parameters of the rotary table through a control system;

[0008] (2) Collect relevant parameters affecting the static pressure turntable through the data acquisition system, such as flow, pressure, temperature and film thickness and the like;

[0009] (3) The collected data are transmitted into a big data calculation system for data classification processing;

[0010] (4) An artificial intelligence system is built by using MATLAB, the processed data are input into an artificial neural network model for training, and the working condition of the static pressure turntable is evaluated and monitored through output values;

[0011] The present application has the following effects:

[0012] 1. The static pressure turntable intelligent monitoring system and method provided by the present application can dynamically monitor the working condition of the static pressure turntable in real time, without the need for maintenance and management by management personnel, and can effectively reduce the operating cost of enterprises;

[0013] 2. The present application can effectively and accurately judge the working condition of the static pressure turntable through the data acquisition system and intelligent algorithm, and the monitoring accuracy is higher as the sample data is more;

[0014] 3. The present application is different from other monitoring systems, and only a single parameter is input for identification and management. The present application is aimed at the influence of multiple factors on the static pressure turntable, and is combined with multiple factors such as film thickness, temperature, pressure and flow for training, so that the working performance and condition of the static pressure turntable can be more comprehensively and accurately understood. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 Intelligent monitoring system diagram.

[0016] Figure 2 Static pressure turntable structure diagram.

[0017] Figure 3 Circular oil pad structure diagram.

[0018] Figure 4 Artificial neural network structure diagram.

[0019] Figure 5 Monitoring algorithm flow chart.

[0020] Figure 6 Artificial intelligence system training result diagram. DETAILED DESCRIPTION

[0021] The present application will be further described in detail below in combination with the drawings and examples.

[0022] The present application implements a static pressure turntable intelligent monitoring system and method, and the implementation of the present application will be specifically described below in combination with the drawings.

[0023] Figure 1The application discloses a static pressure turntable intelligent monitoring system, which mainly comprises a hydraulic pump 8, a data acquisition system and a big data calculation system 7.

[0024] The data acquisition system mainly comprises an oil tank 1, a flow sensor 2, a pressure sensor 3, a displacement sensor 4, a temperature sensor 5, a circular oil pad 6, a rotating disc 9 and an oil pipe 10. The rotating disc 9 is a rotating mechanism of the static pressure turntable, the bottom of the rotating disc 9 is supported by the circular oil pad 6, the circular oil pad 6 is provided with the displacement sensor 4 and the temperature sensor 5, the circular oil pad 6 is supplied with oil by the oil tank 1, the circular oil pad 6 is connected with the oil tank 1 through the oil pipe 10, and the oil pipe 10 is provided with the flow sensor 2 and the pressure sensor 3.

[0025] Figure 2 It is a structural diagram of the static pressure turntable supported by the circular oil pad.

[0026] Figure 3 It is a structural diagram of the circular oil pad, r0 is the radius of the circular oil pad, r1 is the radius of an oil groove, and h is the oil film thickness of the supporting pad.

[0027] The specific steps of the data acquisition system are as follows:

[0028] Step (1), the Reynolds equation form of the circular oil pad is established:

[0029]

[0030] In the formula, p is pressure, h is oil film thickness, U r is radial velocity, is circumferential velocity, U Z is film thickness change rate, and η is viscosity.

[0031] The Reynolds equation is discretized by using a finite difference method, as shown in the following formula:

[0032]

[0033]

[0034] Wherein, is the step length of the r coordinate, is the step length of the θ coordinate. i and j are respectively the numerical count of the r and θ coordinates. A1 i,j , A2 i,j , A3 i,j , A4 i,j , A5 i,j , A6 i,j is a coefficient vector, p i+1,j , p i-1,j , p i,j , p i,j-1 , pi,j+1 represent the node positions after pressure discretization.

[0035] The carrying capacity expression of the oil pad bearing is:

[0036]

[0037] is the dimensionless flow rate, is the dimensionless carrying capacity, P max is the maximum pressure of the oil pump, P r is the pressure set for the rotary table, Q r is the specified oil pump rate, H0 is the initial thickness of the oil film, and R is the radius of the rotating disc.

[0038] Because the dynamic load affects the film thickness, the differential dynamic equilibrium equation of the oil pad is:

[0039]

[0040] where m is the load weight, g is the acceleration of gravity, F(t) is the dynamic load, and t is the time.

[0041] Step (2), establish an artificial neural network training model;

[0042] The BP algorithm is a widely used intelligent algorithm that can be used for classification and prediction. The carrying capacity of the static pressure rotary table is closely related to the film thickness, temperature, pressure, and flow rate. By collecting relevant data through the data acquisition system and inputting them into the big data computing system for processing, the processed data are loaded into the big data computing system 7 for training, and the results based on the static pressure rotary table working performance classification are output, which can well monitor the working state of the rotary table, further improve the quality of the processed parts, and optimize the working performance of the rotary table. The artificial neural network output formula is:

[0043]

[0044] where ω k,j,i is the weight set during the training of the static pressure rotary table, b k,j is the bias set during the training of the static pressure rotary table, fa(z k,j ) is the activation function set during the training of the static pressure rotary table, as shown in Figure 4 The artificial neural network consists of an input layer, two hidden layers, and an output layer.

[0045] The BP algorithm is a multilayer feedforward network trained by the error backpropagation algorithm, and is one of the most widely used neural network models. The training error is:

[0046]

[0047] The calculation formula of the output layer weight thereof is:

[0048]

[0049] The calculation formula of other layer weights is:

[0050]

[0051] The calculation formula of weight update is:

[0052]

[0053] Wherein, λ is a learning rate, and n is an iteration number.

[0054] The convergence criterion of the BP algorithm is:

[0055]

[0056] Step (3), monitoring the static pressure turntable based on the artificial neural network model.

[0057] The program of the artificial neural network model is written by MATLAB, and the calculation process thereof is as shown in Figure 5 The process of monitoring the static pressure turntable based on the artificial neural network model is as follows: (1) first, the sensing unit of the static pressure turntable is controlled through the data acquisition system, the sensing unit includes a flow sensor 2, a pressure sensor 3, a displacement sensor 4 and a temperature sensor 5, the relevant data such as flow, pressure, displacement, temperature and film thickness are collected through the data acquisition system, and then the collected data such as flow, pressure, displacement, temperature and film thickness are divided into a test set and a training set and normalized through a big data calculation system; (2) the normalized data are input to the artificial intelligence system for training, and the classified output value is compared with the target value; (3) if the output value is different from the target value, the related parameters of the static pressure turntable are adjusted in real time to realize the optimal working state of the static pressure turntable.

[0058] In order to more clearly illustrate the method, an example is given below to specifically illustrate the effectiveness of the system and method. As Figure 6 shown, a certain amount of data samples are selected for training, and the consistency of the output value and the target value can be observed, which also shows that the system and method can be used for evaluation and monitoring of the working performance of the turntable.

Claims

1. A hydrostatic rotary table intelligent monitoring system, characterized in that, It includes a hydraulic pump, a data acquisition system, and a big data computing system; the hydraulic pump is connected to the big data computing system through the data acquisition system; the data acquisition system mainly consists of an oil tank, a flow sensor, a pressure sensor, a displacement sensor, a temperature sensor, a circular oil pad, a turntable, and oil pipes; the turntable is the rotating mechanism of the hydrostatic turntable, and its bottom is supported by circular oil pads, on which displacement and temperature sensors are installed; the circular oil pads are supplied with oil from the oil tank, and the circular oil pads are connected to the oil tank through oil pipes, on which flow and pressure sensors are installed; The implementation process of the monitoring system is as follows: (1) Establish a hydrostatic turntable and a circular oil pad model, and use a control system to control each sensing unit to record the real-time parameters of the turntable. (2) Collect relevant parameters affecting the static pressure turntable through the data acquisition system, including flow rate, pressure, temperature and film thickness; (3) The collected data is transmitted to the big data computing system for data classification and processing; (4) Use MATLAB to build an artificial intelligence system, input the processed data into the artificial neural network model for training, and evaluate and monitor the working status of the static pressure rotary table through the output value; The specific steps for setting up a data acquisition system are as follows: Step (1), establish the Reynolds equation form for the circular oil pad: In the above formula, p is the pressure, h is the oil film thickness, and U r For radial velocity, U θ For circumferential velocity, U Z η is the film thickness change rate, and η is the viscosity; Discretize the Reynolds equation using the finite difference method: in, It is the step size of the r-coordinate. The step size is the θ coordinate; i and j are the numerical counts at the r and θ coordinates, respectively; A1 i,j A2 i,j A3 i,j A4 i,j A5 i,j A6 i,j p is the coefficient vector. i+1,j p i-1,j p i,j p i,j-1 p i,j+1 This indicates the node position after pressure discretization; The expression for the load-carrying capacity of an oil-cushion bearing is: For dimensionless flow rate, For dimensionless load-bearing capacity, P max P is the maximum pressure of the oil pump. r The pressure set for the turntable, Q r For the specified oil pump speed, H0 is the initial thickness of the oil film, and R is the radius of the turntable; Dynamic load affects film thickness; the differential dynamic equilibrium equation for the oil pad is: Where m is the loaded weight, g is the gravitational acceleration, F(t) is the dynamic load, and t is the time; Step (2): Establish an artificial neural network training model; The processed data is loaded into a big data computing system for training, and the results based on the performance classification of the hydrostatic rotary table are output. The working status of the rotary table is monitored, the quality of processed parts is improved, and the working performance of the rotary table is optimized. Step (3): Monitor the hydrostatic turntable based on an artificial neural network model; The program for the artificial neural network model is written in MATLAB. The process of monitoring the static pressure turntable based on the artificial neural network model is as follows: (1) First, the sensing unit of the static pressure turntable is controlled by the data acquisition system. The sensing unit includes a flow sensor, a pressure sensor, a displacement sensor and a temperature sensor. The flow rate, pressure, displacement, temperature and film thickness are collected by the data acquisition system. Then, the collected flow rate, pressure, displacement, temperature and film thickness are divided into a test set and a training set and normalized by the big data computing system. (2) The normalized data is input into the artificial intelligence system for training and the output value after classification is compared with the target value. (3) If the output value is different from the target value, the relevant parameters of the static pressure turntable are adjusted in real time to achieve the optimal working state of the static pressure turntable.

Images