Farmland irrigation water pump group automatic control system and method

Abstract

The invention belongs to the technical field of farmland irrigation, and discloses an automatic control system and method for a farmland irrigation water pump group, the system comprises a master control platform and a plurality of water pump devices in a farmland, and each device is provided with a PLC (Programmable Logic Controller) and a sensor. The control platform comprises a data acquisition module, a humidity detection module and a water pump state detection module, humidity data are fused through the D-S theory, the opening degree of a water pump ball valve is controlled, and vibration and current data are analyzed through an RBF neural network model for fault diagnosis. According to the technical scheme, the purpose of unattended automatic shutdown of the farmland irrigation water pump group equipment can be achieved, and the production efficiency of farmland irrigation is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of farmland irrigation, and in particular relates to an automatic control system and method for a farmland irrigation water pump group. Background Art

[0002] Organic dryland farming is a traditional technical feature of agriculture. We should adhere to the path of organic dryland farming and improve the technical system of organic dryland farming, which provides a direction for the development of modern agriculture in Shanxi under the new situation. Therefore, farmland irrigation has become a key link in the process of agricultural planting and production, and plays a decisive role in increasing crop yields.

[0003] However, there are still some problems with existing agricultural irrigation technology: (1) It is still based on traditional production experience and lacks precise control methods. It is difficult to ensure the accuracy of farmland irrigation, resulting in serious waste of water resources; (2) Irrigation technology is backward, monitoring and control are all managed manually, lacking technological innovation, and has disadvantages such as high labor intensity and poor human-computer interaction capabilities, which seriously affect the quality and yield of crops. Summary of the invention

[0004] The purpose of the present invention is to provide an automatic control system and method for a farmland irrigation water pump group to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above-mentioned object, the present invention provides an automatic control system for a farm irrigation water pump group, comprising a general control platform and a plurality of water pump devices connected to the general control platform in communication, each of the water pump devices is installed in the farmland, and each of the water pump devices is installed with a PLC controller, and each of the PLC controllers is connected with a plurality of sensors;

[0006] The control platform comprises a data acquisition module, a humidity detection module and a water pump operation status detection module which are connected in sequence; the data acquisition module is in communication connection with each of the PLC controllers.

[0007] Optionally, the sensor includes a vibration sensor, a current sensor and a humidity sensor, the vibration sensor and the current sensor are installed on corresponding water pump equipment, and the humidity sensor is installed in the farmland.

[0008] Optionally, the data acquisition module is connected to each of the PLC controllers via a VPN network communication.

[0009] A method for automatically controlling a farmland irrigation water pump group, applied to the automatic control system of a farmland irrigation water pump group, comprising:

[0010] Collect real-time humidity monitoring data of farmland through humidity sensors;

[0011] The real-time humidity monitoring data is sent to the data acquisition module through the PLC controller, and the data acquisition module transmits the received real-time humidity monitoring data to the humidity detection module, and the humidity detection module performs data fusion on the real-time humidity monitoring data based on the DS evidence theory to obtain a comprehensive humidity value;

[0012] Based on the comprehensive humidity value, controlling the ball valve opening of the water pump device;

[0013] In the process of controlling the water pump device, real-time characteristic signal data of the water pump device is obtained, wherein the real-time characteristic signal data includes vibration data and current data;

[0014] The real-time characteristic signal data is input into a water pump fault identification model in a water pump operation status detection module to perform fault detection, obtain a fault diagnosis result, and determine whether the water pump equipment has a fault based on the fault diagnosis result; wherein the water pump fault identification model is constructed based on a radial basis function neural network.

[0015] Optionally, the data fusion of the real-time humidity monitoring data based on the DS evidence theory specifically includes:

[0016] Performing basic probability assignment on the humidity monitoring data acquired by each humidity sensor, and constructing a recognition framework;

[0017] The basic probability assignments of different humidity sensors are fused based on the Dempster combination rule to obtain the comprehensive humidity value.

[0018] Optionally, the training process of the water pump fault identification model specifically includes:

[0019] Acquire training data, wherein the training data includes characteristic signal training data of the water pump equipment and corresponding fault results;

[0020] An initial water pump fault identification model is constructed, the training data is input into the initial water pump fault identification model for fault detection, and training is performed with the goal of minimizing the loss between the initial training result after fault detection and the fault result corresponding to the characteristic signal training data to obtain a trained water pump fault identification model.

[0021] The technical effects of the present invention are:

[0022] The present invention takes farmland irrigation water pump group equipment as the object, adjusts the water pump state accordingly according to the humidity value of the farmland, realizes the goal of unattended automatic shutdown of the farmland irrigation water pump group equipment, and improves the production efficiency of farmland irrigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0024] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0025] Figure 1 is a diagram of a sensor-level fusion structure in an embodiment of the present invention;

[0026] Figure 2 Schematic diagram of the radial basis neural network structure in an embodiment of the present invention;

[0027] Figure 3 It is an implementation flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0028] Various exemplary embodiments of the present invention will now be described in detail. This detailed description should not be considered as limiting the present invention, but should be understood as a more detailed description of certain aspects, features, and embodiments of the present invention.

[0029] It should be understood that the terms described in the present invention are only for describing special embodiments and are not intended to limit the present invention. In addition, for the numerical range in the present invention, it should be understood that each intermediate value between the upper and lower limits of the scope is also specifically disclosed. Each smaller range between the intermediate value in any stated value or stated range and any other stated value or intermediate value in the described range is also included in the present invention. The upper and lower limits of these smaller ranges can be independently included or excluded in the scope.

[0030] It will be apparent to those skilled in the art that various modifications and variations may be made to the specific embodiments of the present invention description without departing from the scope or spirit of the present invention. Other embodiments derived from the present invention description will be apparent to those skilled in the art. The present application description and examples are exemplary only.

[0031] The words “include,” “including,” “have,” “contain,” etc. used in this article are open-ended terms, meaning including but not limited to.

[0032] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0033] Embodiment 1

[0034] like Figure 1 - Figure 3 As shown, in this embodiment, an automatic control system for a farm irrigation water pump group is provided, including a general control platform and a plurality of water pump devices connected to the general control platform for communication, each of the water pump devices is installed in the farmland, and each of the water pump devices is installed with a PLC controller, and each of the PLC controllers is connected with a plurality of sensors;

[0035] The control platform comprises a data acquisition module, a humidity detection module and a water pump operation status detection module which are connected in sequence; the data acquisition module is in communication connection with each of the PLC controllers.

[0036] A method for automatically controlling a farmland irrigation water pump group, applied to the automatic control system of a farmland irrigation water pump group, comprising:

[0037] Collect real-time humidity monitoring data of farmland through humidity sensors;

[0038] The real-time humidity monitoring data is sent to the data acquisition module through the PLC controller, and the data acquisition module transmits the received real-time humidity monitoring data to the humidity detection module, and the humidity detection module performs data fusion on the real-time humidity monitoring data based on the DS evidence theory to obtain a comprehensive humidity value;

[0039] Based on the comprehensive humidity value, controlling the ball valve opening of the water pump device;

[0040] In the process of controlling the water pump device, real-time characteristic signal data of the water pump device is obtained, wherein the real-time characteristic signal data includes vibration data and current data;

[0041] The real-time characteristic signal data is input into a water pump fault identification model in a water pump operation status detection module to perform fault detection, obtain a fault diagnosis result, and determine whether the water pump equipment has a fault based on the fault diagnosis result; wherein the water pump fault identification model is constructed based on a radial basis function neural network.

[0042] The present application is based on a monitoring and surveillance system for a farmland irrigation water pump group, and comprehensively utilizes a variety of new technologies such as computer technology, microprocessor technology, automatic control technology, database technology, network technology, digital electronic technology, and signal processing. It takes a farmland irrigation water pump group equipment as the object, and adjusts the speed of the motor accordingly according to the humidity value of the farmland. The motor drives the water pump to pump water from the water source. When irrigation is needed, the solenoid valve automatically opens, and water is supplied to the sprinkler through the main pipe and the branch pipe. The sprinkler automatically rotates at its respective rotation angle, thereby achieving the goal of unattended automatic shutdown of the farmland irrigation water pump group equipment, thereby improving the production efficiency of farmland irrigation.

[0043] 1. The overall framework of this embodiment: The automatic control system of farm irrigation water pump group, because it is necessary to realize the centralized, remote monitoring and control of multiple areas, multiple pieces and multiple places, this system has a three-layer network architecture: perception layer, network layer, and application layer. The perception layer uses sensor node units to obtain real-time situation data of farmland, providing the most direct data for automatic control and early warning; the network layer uses field bus and VPN network data transmission channels to complete reliable data transmission; the application layer is based on analysis, control and early warning platform. This embodiment introduces the automatic control system of water pump group based on PowerBuilder development platform.

[0044] 2. Information perception process: Information perception is the perception layer of the three-layer network architecture of the system in this embodiment. It uses sensor node units to obtain real-time situation data of farmland and provide data basis for automatic control and early warning. The information collected by this system includes the humidity of farmland, the current of water pump motor, and the vibration signal of water pump shaft end. This information is collected by the front-end real-time link and transmitted to the back-end development platform, which is analyzed by the back-end software and gives control commands.

[0045] The background development platform is PowerBuilder, and the real-time control core of each water pump is PLC. Since the farmland irrigation water pump group includes water pump equipment in multiple zones, multiple pieces and multiple locations, the platform is connected to more PLC devices, and data transmission and communication are carried out through the VPN network.

[0046] The information humidity, water pump motor current, and water pump shaft end vibration sensor of this embodiment are configured as follows: the vibration sensor is an acceleration sensor, which is installed at the water pump drive end. By analyzing the signal spectrum of the vibration sensor, the working state of the water pump can be known, and whether the water pump is faulty can be judged. The current is measured by the Hall current sensor, and the magnitude of the current can be used to know whether the water pump motor is working. The humidity of the farmland is measured by the humidity sensor. 3x3=9 sensors are placed per acre of land. Whether to start the pump for irrigation is determined based on the sensor data. The above signals are collected by the PLC and transmitted to the background computer.

[0047] 3. Signal communication: Signal communication is the network layer of the three-layer network architecture of the system in this embodiment. It uses field bus and VPN network data transmission channels to complete reliable data transmission.

[0048] Each PLC device has an IP address. In this embodiment, data communication uses Winsock three-party controls in the PowerBuilder development platform. The background host computer is the client, and the real-time control core PLC is the server. As many Winsocks are needed as there are PLCs. Winsock is initialized as a client in the Open event of the PowerBuilder window.

[0049] 4. Intelligent processing: Since the present embodiment involves a large number of water pump equipment and sensors, it is necessary to group and number the irrigated fields, field-installed sensors, and water pumps used for irrigation so as to carry out targeted monitoring and control.

[0050] The irrigated field is represented by td, and the field installation sensor information is represented by sd (humidity signal), i (water pump motor current signal), and zd (vibration signal). Then the values ​​of the pth irrigated field are represented by td p 、sd pk (k=1~9),i p 、zd p In this embodiment, these signals are analyzed by expert rules to give water pump start and stop control commands; and the running status of the water pump is monitored by a neural network.

[0051] (1) Automatic control: The humidity unit of the humidity sensor is %RH, and the measurement range is 0-100%RH. For the pth irrigated field, since it is not very flat, it is difficult to obtain the true comprehensive humidity by simply relying on the simple average or weighted calculation of a single or several humidity sensors. It is necessary to integrate the data of all installed humidity sensors.

[0052] DS evidence theory is to gradually narrow down the hypothesis set through the accumulation of evidence. A basic strategy of evidence theory is to divide the evidence set into two or more unrelated parts, and use them to independently judge the identification framework, and then combine them using the Dempster combination rule. The DS evidence combination rule is the core content of evidence theory. It is a method to calculate the comprehensive impact of multiple evidences on the hypothesis in the process of evidence accumulation, that is, the comprehensive degree of confidence in the hypothesis under the influence of multiple evidences. More specifically, it is a mathematical means to integrate multiple evidences from multiple angles for information fusion on the same problem.

[0053] Therefore, the fusion rule of DS evidence theory is defined as: Assume m1, m2, ..., m nis the basic probability assignment function derived from n independent evidence sources on the same identification framework Θ, then the combination rule can be expressed as:

[0054]

[0055] In the above formula, k is the conflict coefficient, which is defined as:

[0056]

[0057] This embodiment is based on the information of multiple humidity sensors and is fused at the sensor level. Before fusion, the humidity values ​​of each humidity sensor are used. The category of its output state is used as the identification framework of DS evidence theory, and the humidity value of each humidity sensor is assigned as the basic probability. The humidity values ​​of each humidity sensor are fused using the evidence combination formula in DS evidence theory to obtain the final humidity of the irrigated field, such as Figure 1 shown.

[0058] In this way, when the water pump is in normal state, when the humidity fusion value is 0 <sd p ≤20%RH, the water pump responsible for the field is started for irrigation, and the opening of the control ball valve is 100%; when the humidity fusion value is 20 <sd p ≤40%RH, the water pump responsible for the field is started for irrigation, and the opening of the control ball valve is 80%; when the humidity fusion value is 40 <sd p ≤60%RH, the water pump responsible for the field is started to irrigate, and the opening of the control ball valve is 60%; when the humidity fusion value is 60 <sd p ≤80%RH, the water pump responsible for the field is started for irrigation, and the opening of the control ball valve is 40%; when the humidity fusion value is 80 <sd p ≤95%RH, the water pump responsible for the field is started for irrigation, and the opening of the ball valve is controlled to be 20%; when the humidity fusion value sd p >95%RH, the control ball valve is closed and the water pump responsible for this field is stopped.

[0059] In this embodiment, the sensor information of the pth field is used to define the recognition framework U as {0 <sd p ≤20%RH for X1,20 <sd p ≤40%RH for X2,40 <sd p ≤60%RH for X3,60 <sd p ≤80%RH for X4,80 <sd p ≤95%RH for X5, sd p >95%RH is X6}, and the test is continued for 100 times. The data of identification frame U obtained in the experiment are as follows:

[0060] m1(X1)=0.074, m1(X2)=0.068, m1(X3)=0.126, m1(X4)=0.061, m1(X5)=0.317, m1(X6)=0.065;

[0061] m2(X1)=0.079, m2(X2)=0.126, m2(X3)=0.074, m2(X4)=0.091, m2(X5)=0.336, m2(X6)=0.031;

[0062] m3(X1)=0.073, m3(X2)=0.052, m3(X3)=0.043, m3(X4)=0.031, m3(X5)=0.417, m3(X6)=0.112;

[0063] m4(X1)=0.043, m4(X2)=0.132, m4(X3)=0.015, m4(X4)=0.041, m4(X5)=0.518, m4(X6)=0.055;

[0064] m5(X1)=0.024, m5(X2)=0.053, m5(X3)=0.106, m5(X4)=0.075, m5(X5)=0.445, m5(X6)=0.041;

[0065] The basic credibility distribution of the five bodies of evidence is shown in Table 1;

[0066] Table 1 Basic credibility distribution of 5 evidence bodies

[0067]

[0068] According to DS theory: Bel(X j )=m i (X j ), pl(X j )=Bel(X j )+m i (θ), we get the data shown in Table 2. From the results, we can see that no conclusion can be drawn from evidence 1 and evidence 2, and no definite conclusion can be drawn from evidence 3 and evidence 5 either.

[0069] Table 2 Reliability intervals and diagnostic results of each body of evidence acting alone

[0070]

[0071]

[0072] Through the DS evidence fusion method, each piece of evidence is fused separately. In this embodiment, evidence 1 and evidence 2 are fused, evidence 1, evidence 2 and evidence 3 are fused, evidence 1, evidence 2, evidence 3 and evidence 4 are fused, and evidence 1, evidence 2, evidence 3, evidence 4 and evidence 5 are fused. All of these can well obtain deterministic conclusions.

[0073] The following describes the calculation method for fusing three pieces of evidence. According to the above formulas (1) and (2), if evidence 1, evidence 2, and evidence 3 are fused, the conflict coefficient is calculated first:

[0074]

[0075]

[0076] Through calculation, we can know that even if only the fusion of evidence 1 and 2 is used, the conclusion can be drawn well. Therefore, Table 3 only lists the fusion results of evidence 1 and 2, and evidence 1, 2 and 3.

[0077] Table 3 Confidence intervals and diagnostic results after fusion of multiple evidence bodies

[0078]

[0079] The humidity fusion algorithm of this embodiment is developed on the PowerBuilder front-end platform.

[0080] (2) Condition Monitoring: This embodiment uses artificial intelligence technology based on radial basis function (RBF) neural network to monitor the health status of water pump equipment, and is developed with PowerBuilder as the front-end platform.

[0081] The radial basis function (RBF) neural network is a regularized network with three layers: input layer, intermediate layer and output layer. It has only one hidden layer, and the hidden unit is the basis function Φ(x, x i ), in this embodiment, Φ(x, x i )Select Gaussian function:

[0082]

[0083] As a basis function, where p is the dimension of the input vector, that is, the number of inputs, the output of RBF is:

[0084]

[0085] Among them, X i ,σ,ω are the center, function width and weight of the hidden unit respectively, as Figure 2 shown.

[0086] In this way, the training of the radial basis neural network includes the training of the output unit weights ω, the center X of the hidden unit i In this embodiment, the 8-layer wavelet packet of the vibration signal S0 (0Hz-19Hz), S1 (19Hz-39Hz), S4 (78Hz-98Hz), S5 (98Hz-118Hz), S6 (118Hz-138Hz), S 10 (196Hz-215Hz), S 11 (215Hz-235Hz), S 12 (235Hz-255Hz), S 13 (255Hz-275Hz), S 14 (275Hz-294Hz), S 15 (294Hz-313Hz), S 16 (313Hz-333Hz), S 17 (333Hz-353Hz), S 18 (353Hz-373Hz), S 19 (373Hz-393Hz), S 20 (393Hz-412Hz), S 21 (412Hz-431Hz), S 22 (431Hz-451Hz), S 30 (588Hz-608Hz), S 31 (608Hz-627Hz), S 32 (627Hz-647Hz), S 33 (647Hz-666Hz), S 34 (666Hz-686Hz), S 35 (686Hz-706Hz), S 36 (706Hz-725Hz), S 37 (725Hz-745Hz), S 38 (745Hz-764Hz), S 40 The energy values ​​of these 28 frequency bands (784Hz-804Hz) are taken as fault characteristic values, and S3 (29Hz-39Hz), S4 (39Hz-49Hz), S5 (49Hz-59Hz), S6 (59Hz-68Hz), S7 (68Hz-78Hz), S8 (78Hz-88Hz), S9 (88Hz-98Hz), S10 (101Hz-110Hz), S11 (101Hz-110Hz), S12 (101Hz-110Hz), S13 (101Hz-110Hz), S14 (101Hz-110Hz), S15 (101Hz-110Hz), S16 (101Hz-110Hz), S17 (101Hz-110Hz), S18 (101Hz-110Hz), S19 (101Hz-110Hz), S20 (101Hz 10 (98Hz-108Hz), S 11 (108Hz-117Hz), S 15(147Hz-156Hz), S 16 (156Hz-166Hz), S 17 (166Hz-176Hz), S 18 (176Hz-186Hz), S 19 (186Hz-195Hz), S 25 (245Hz-254Hz), S 35 (342Hz-352Hz), S 56 (548Hz-558Hz), S 66 (646Hz-656Hz) 18 frequency band energy values ​​are used as fault characteristic values, with a total of 46 characteristic quantities, namely p = 46, to evaluate 9 types of conditions, including water pump motor bearing, squirrel cage, stator and air gap eccentricity, water pump impeller mechanism failure, centrifugal pump water discharge, water pump cavitation, foundation looseness, and water pump normality. These 46 characteristic signals are used as input X (x1, x2, ..., x 45 , x 46 ), the value of the 9-category state, that is, N = 9, is used as the center X of the RBF i (x i1 , x i2 , ..., x i45 , x i46 ), output unit weights ω(ω1, ω2, …, ω8, ω9) and widths σ(σ1, σ2, …, σ 45 , σ 46 ) is directly calculated by the least square method. In this way, a radial basis neural network with 46 inputs, 9 hidden units and 1 output is obtained. The output is the running state of the water pump, including normal state, water pump motor bearing ball fault, inner ring fault, outer ring fault, squirrel cage broken bar fault, air gap eccentricity fault, water pump impeller mechanism fault, centrifugal pump water failure fault, water pump cavitation fault, foundation loosening fault and other states (0 represents normal, 1, 2, 3, 4, 5, 6, 7, 8, 9 represent the nine fault states of the water pump).

[0087] After the implementation of the scheme of this embodiment, real-time and effective transmission of multi-sensor information in the field is realized; online monitoring of operating parameters and equipment status is realized, and real-time monitoring, analysis and control of the operating status of the water pump system are realized; coordinated control of water pump equipment groups in multiple zones, multiple areas and multiple locations is achieved.

[0088] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An automatic control system for a farm irrigation water pump group, characterized in that: It includes a general control platform and several water pump devices that are connected to the general control platform in communication, each of the water pump devices is installed in the farmland, and each of the water pump devices is installed with a PLC controller, and each of the PLC controllers is correspondingly connected with several sensors; The control platform comprises a data acquisition module, a humidity detection module and a water pump operation status detection module which are connected in sequence; the data acquisition module is in communication connection with each of the PLC controllers.

2. The automatic control system for farmland irrigation water pump group according to claim 1 is characterized in that: The sensors include a vibration sensor, a current sensor and a humidity sensor. The vibration sensor and the current sensor are installed on corresponding water pump equipment, and the humidity sensor is installed in the farmland.

3. The automatic control system for farmland irrigation water pump group according to claim 1 is characterized in that: The data acquisition module is connected to each of the PLC controllers via a VPN network communication.

4. A method for automatically controlling a farm irrigation water pump group, applied to an automatic control system for a farm irrigation water pump group as claimed in any one of claims 1 to 3, characterized in that: include: Collect real-time humidity monitoring data of farmland through humidity sensors; The real-time humidity monitoring data is sent to the data acquisition module through the PLC controller, and the data acquisition module transmits the received real-time humidity monitoring data to the humidity detection module, and the humidity detection module performs data fusion on the real-time humidity monitoring data based on the DS evidence theory to obtain a comprehensive humidity value; Based on the comprehensive humidity value, controlling the ball valve opening of the water pump device; In the process of controlling the water pump device, real-time characteristic signal data of the water pump device is obtained, wherein the real-time characteristic signal data includes vibration data and current data; The real-time characteristic signal data is input into a water pump fault identification model in a water pump operation status detection module to perform fault detection, obtain a fault diagnosis result, and determine whether the water pump equipment has a fault based on the fault diagnosis result; wherein the water pump fault identification model is constructed based on a radial basis function neural network.

5. The automatic control method of a farm irrigation water pump group according to claim 4, characterized in that: The data fusion of the real-time humidity monitoring data based on the DS evidence theory specifically includes: Performing basic probability assignment on the humidity monitoring data acquired by each humidity sensor, and constructing a recognition framework; The basic probability assignments of different humidity sensors are fused based on the Dempster combination rule to obtain the comprehensive humidity value.

6. The automatic control method of a farm irrigation water pump group according to claim 4, characterized in that: The training process of the water pump fault identification model specifically includes: Acquire training data, wherein the training data includes characteristic signal training data of the water pump equipment and corresponding fault results; An initial water pump fault identification model is constructed, the training data is input into the initial water pump fault identification model for fault detection, and training is performed with the goal of minimizing the loss between the initial training result after fault detection and the fault result corresponding to the characteristic signal training data to obtain a trained water pump fault identification model.

Images