Water film generating device for pneumatic material conveying system and intelligent control method of water film generating device

Abstract

The invention relates to the technical field of pneumatic conveying, in particular to a water film generation device for a pneumatic material conveying system and an intelligent control method thereof, and the water film generation device comprises a water film generation device, a PID controller and a recurrent neural network; the water film generating device is used for providing high-pressure water for the nozzle, water mist is formed through the nozzle and is uniformly sprayed into a pneumatic material conveying pipeline, and a water film is formed in the pipeline; the intelligent control method is matched with a monitoring device of the pneumatic material conveying system to set initial parameters, the air saturation water content is calculated, the PID controller is used for regulating and controlling the water injection rate, training and learning are conducted on the pressure, temperature, humidity and air speed in the pneumatic material conveying process based on the recurrent neural network, and the air saturation water content is calculated. Initial parameters are corrected in real time according to the predicted air saturation water content and pressure loss, so that the material conveying efficiency and stability are improved, energy consumption and material loss are reduced, meanwhile, a pipeline is cooled, and the phenomena of dust and wire drawing caused by heat generated by material friction are prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of pneumatic conveying, and in particular to a water film generating device for a pneumatic material conveying system and an intelligent control method thereof. Background Art

[0002] Pneumatic conveying is a green transportation method that uses airflow to transport powder (granular) materials in a relatively closed pipeline along the airflow direction. It is widely used in agriculture, chemical industry, food processing, machinery manufacturing, medicine and other fields. The closure of the pneumatic conveying process can prevent the material from being contaminated by the outside world, and can flexibly change the conveying route by multi-point feeding and multi-point unloading. At the same time, the conveying efficiency of powder and granular materials is also high. However, pneumatic conveying also has many problems, such as pipeline wear and wire drawing. In the pneumatic conveying system, due to the interaction between particles and particles, particles and walls, material particles will inevitably produce friction and damage, resulting in dust and wire drawing in the pipe, which has an adverse effect on the stability and continuity of particle transportation, and also aggravates energy consumption loss. Another common material conveying method is hydraulic conveying. It uses the flow of water to transport materials, with low energy consumption, high efficiency, and no dust and wire drawing due to frictional heat generation, but the conveying distance is limited and it is necessary to extract water from the water and material mixture.

[0003] From the above, it can be seen that wire drawing and wear occur during pneumatic conveying, which affects the stability of the conveying system and increases energy consumption. Therefore, referring to hydraulic conveying, the pneumatic conveying method is improved and upgraded. However, the current common means of data collection and processing during the conveying process are inefficient, with low precision and accuracy, and cannot effectively realize automated decision-making for continuous dynamic changes in data. Therefore, a more intelligent means and method can be used to perform feature processing and data analysis and prediction on process data, optimize the overall conveying efficiency and stability of pneumatic conveying, improve precision and accuracy, and reduce energy consumption and material loss.

[0004] In order to solve the above problems, the present invention proposes a water film generating device for a pneumatic material conveying system and an intelligent control method thereof, wherein the water film generating device is used to provide high-pressure water to a nozzle, and water mist is formed through the nozzle and evenly sprayed into a pipe for pneumatic material conveying, and a water film is formed in the pipe; the intelligent control method cooperates with a monitoring device of the pneumatic material conveying system to set initial parameters, calculate the saturated water content of the air, and use a PID controller to adjust the water injection flow rate, and trains and learns the pressure, temperature, humidity and air velocity of the pneumatic material conveying process based on a recurrent neural network, and corrects the initial parameters in real time according to the predicted saturated water content of the air and pressure loss, thereby improving the material conveying efficiency and stability, reducing energy consumption and material loss, and cooling the pipe at the same time to prevent the friction heat of the material from generating dust and wire drawing. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the defects in the prior art, thereby providing a water film generating device for a pneumatic material conveying system and an intelligent control method thereof, including an air intake port, a third electromagnetic valve, a first air intake pipe, a water tank, an electromagnetic flowmeter, a first electromagnetic valve, an intermediate pipe, a nozzle pipe, a nozzle, a second electromagnetic valve, a second air intake pipe, a temperature and humidity sensor, a PID controller, and a recurrent neural network. The air intake port is used for the inlet of compressed gas for the entire air film generating device; the third electromagnetic valve is used to control the on and off of the air flow; the first air intake pipe is used to introduce the compression of the air film generating device into the water tank, and is used to balance the water tank and the pressure; the water tank is used to provide a water source for the air film generating device; the electromagnetic flowmeter is used to measure the flow rate of water in the air film generating device in real time and accurately; the first electromagnetic valve is used to accurately adjust the flow rate of water entering the system; the intermediate pipe is used to transport the high-pressure water flow of the air film device; the nozzle pipe is used to divert the high-pressure water flow; the nozzle is used to generate spray for the air film device; the second electromagnetic The valve is used to open and close the compressed gas; the second air inlet pipe is used to directly transport the compressed air to the intermediate pipe; the temperature and humidity sensor is used to detect the temperature and humidity in the material boot and the front and rear end pipes of the conveying; the PID controller is used to adjust the water injection amount, and the recurrent neural network is trained and learned about the pressure, temperature, humidity and air velocity of the pneumatic material conveying process, and the initial parameters are corrected in real time according to the predicted air saturated water content and pressure loss, so as to improve the precision and accuracy of material conveying, reduce energy consumption and material loss, and cool the pipeline at the same time to prevent the friction heat of the material from generating dust and wire drawing.

[0006] The technical solution specifically adopted by the present invention is:

[0007] A water film generating device for a pneumatic material conveying system and an intelligent control method thereof, characterized in that: it includes an air inlet port, a third solenoid valve, a first air inlet pipe, a water tank, a first solenoid valve, an intermediate pipe, a nozzle pipe and a nozzle; the input end of the first air inlet pipe is connected to the air inlet port for compressed gas input, the third solenoid valve is arranged on the first air inlet pipe, the output end of the first air inlet pipe is inserted into the water tank input port for balancing the water pressure in the water tank, the water tank output port is connected to the intermediate pipe input end through the first solenoid valve, the first solenoid valve is used to adjust the water injection flow, the intermediate pipe output end is connected to the nozzle through the nozzle pipe, and the nozzle is used to provide a water film for the pneumatic material conveying pipe.

[0008] The water film generating device is characterized in that it also includes a second air intake pipe, on which a second electromagnetic valve is provided, the input end of the second air intake pipe is connected to the first air intake pipe near the air intake port, and the output end of the second air intake pipe is connected to the intermediate pipe.

[0009] The water film generating device is characterized in that when the second electromagnetic valve is opened alone, compressed gas is input into the nozzle for cleaning the pneumatic material conveying pipeline.

[0010] The water film generating device is characterized in that it also includes an electromagnetic flowmeter for recording the water flow size, the electromagnetic flowmeter is connected to the water tank through a flange, and the electromagnetic flowmeter is connected to the first electromagnetic valve through a flange.

[0011] The water film generating device is characterized in that the nozzle pipeline is at least two pipelines, the nozzle is at least two nozzles, and the nozzle pipelines are respectively connected to the corresponding nozzles.

[0012] The water film generating device is characterized in that: the nozzle is an annular nozzle, which includes an annular pipe and a Venturi tube connected in sequence, the Venturi tube is multiple and evenly arranged on the annular pipe, the Venturi tube is arranged to form an angle of 45° with the horizontal plane, and the Venturi tube includes a contraction tube, a throat tube and a diffusion tube connected in sequence, and the throat tube has the smallest diameter among the three.

[0013] An intelligent control method for the water content of air in a pneumatic material conveying pipeline uses a PID controller to adjust the first electromagnetic valve to adjust the water injection flow in real time. The PID controller is a very classic controller, which consists of three parts: proportional (P), integral (I) and differential (D). It can control the system deviation, cumulative deviation and deviation change rate respectively. Its expression is:

[0014]

[0015] Where u(t) is the control signal, e(t) is the error signal, and K p , K i , K d It is a parameter of the PID controller. If the monitored pressure loss is lower than the pressure loss corresponding to the last system parameter setting value, the water injection amount is fine-tuned positively, otherwise it is reduced. By adjusting the response control signal, the water injection amount of the pneumatic conveying process is dynamically monitored and adjusted, thereby improving the operating efficiency and stability of the system.

[0016] An intelligent control method for air moisture content in a pneumatic material conveying pipeline uses a recurrent neural network algorithm to construct a prediction model for air saturation moisture content and pressure loss during pneumatic material conveying. The algorithm implementation process is as follows: a recurrent neural network is a method for (1) ,…,x (t) A neural network model developed for sequence data. The recurrent neural network needs to target x (1) ,…,x (t)Sequence data has a certain logical association, that is, the state at time t that needs to be obtained only depends on the past x (1) ,…,x (t－1) and the input x at time t t The predicted value y(t) at time t depends on the previous x. (1) ,…,x (t－1) The corresponding y (1) ,…,y (t－1) and the input x at time t t Therefore, the present invention chooses to use a three-layer perceptual recurrent neural network to construct a prediction model for air saturated moisture content and pressure loss during pneumatic conveying, which includes an input layer, a hidden layer and an output layer. Neurons in the same layer are not connected to each other, while neurons between two adjacent layers are fully connected to each other. Pressure P, temperature T, humidity AH and air velocity v are used as input layers x1, x2, x3, x4, the predicted target air saturated moisture content is used as output layer y1, and the predicted pressure loss is used as output layer y2. After determining the input layer and output layer, in the forward propagation stage, the input layer obtains the input signal, that is, all influencing factors x1,…,x1 that are interrelated with air saturated moisture content and pressure loss. n , and pass it to each neuron in the hidden layer. Then, the hidden layer processes these signals and passes the processing results to the output layer. The back propagation algorithm (BP algorithm) is used to iterate and fit the data, that is, to solve the local extreme value by gradient descent, for each influencing factor x1, ..., x n , each neuron in the hidden layer has a corresponding input and output signal, marked as u j and h j , these two signals can be calculated by the following formulas:

[0017]

[0018] w ij is the weight between input layer neuron i and hidden layer neuron j; θ j is the random disturbance term,

[0019] The input signal lk and output signal Ck of the output layer can be calculated by the following formulas:

[0020]

[0021] Where: V jk is the weight between hidden layer neuron j and output layer neuron k; γ k is a random disturbance term; the output signal C of the output layer k The required prediction target is the air saturated moisture content; the input signal l k is the output signal h corresponding to each neuron in the hidden layerj The linear combination of the forward propagation output signal C k The difference from the true signal is the deviation. The deviation of each output layer neuron dk and hidden layer neuron deviation can be calculated by the following formula:

[0022] d k =(y k -C k )C k (1-C k ),k=1 (6)

[0023]

[0024] The bias is fed back from the output layer to the hidden layer, and the connection weights between the output layer and each hidden neuron are updated using the following formula (8). The connection weights between each hidden layer neuron and each influencing factor x1, ..., x2 in the input layer are updated using the following formula (8). n The weight between .

[0025]

[0026] Where: i = 1, 2, ..., n; j = 1, 2, 3, 4; k = 1; a1 and a2 are learning rates ranging from 0 to 1. The trained model is deployed in the actual pneumatic conveying system for real-time monitoring, prediction and optimization control.

[0027] A water film generating device for a pneumatic material conveying system and an intelligent control method thereof, characterized in that: a monitoring device of the pneumatic material conveying system is taken as the core, and the specific steps are as follows:

[0028] (1) Initially set the air flow rate and particle feeding rate;

[0029] (2) Collect the pressure, temperature, humidity and air velocity before and after the material shoe and before and after the transportation;

[0030] (3) Calculate the saturated water content of air;

[0031] (4) Set the injection flow rate based on the calculated air saturation water content.

[0032] (5) High-pressure spraying based on water injection flow rate;

[0033] (6) Determine whether the system pressure loss has decreased?

[0034] (7) If not, use PID controller to fine-tune and reduce the water injection amount;

[0035] (8) If yes, use PID controller to fine-tune and increase the water injection amount;

[0036] (9) Record the adjusted water injection volume, pressure, temperature, humidity and gas velocity.

[0037] (10) Inputting the recurrent neural network for training, predicting the output air saturation moisture content and pressure loss, and correcting the initial air flow rate and particle descent rate;

[0039] (11) Determine whether the delivery is completed?

[0040] (12) If not, return to step (2);

[0041] (13) If yes, flush the pipes and nozzles of the water film generating device;

[0042] (12)End.

[0043] The technical solution of the present invention has the following advantages:

[0044] (1) A uniform water mist is generated in the pneumatic material conveying system based on a water film generating device to cool the conveyed materials; (2) The pneumatic material conveying monitoring system is used to collect relevant parameters, calculate the air saturated water content, and set the water injection flow rate. The PID controller is used to control the water injection amount in real time. The pressure, temperature, humidity and air velocity of the pneumatic material conveying process are trained and learned in combination with the recurrent neural network algorithm. The initial air flow rate and particle descent rate are corrected in real time according to the predicted air saturated water content and pressure loss, so as to improve the overall conveying efficiency and stability of the pneumatic conveying, reduce energy consumption and material loss, and prevent the frictional heat of the material from generating dust and wire drawing. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 This is a schematic structural diagram of a water film generating device for a pneumatic material conveying system according to an embodiment of the present invention;

[0047] Figure 2 Schematic diagram of the nozzle structure;

[0048] Figure 3 Schematic diagram of a recurrent neural network;

[0049] Figure 4 This is a flow chart of the working steps of a method for intelligently controlling the moisture content of air in a pneumatic material conveying pipeline according to an embodiment of the present invention.

[0050] Description of Reference Numerals

[0051] 1-first air intake pipe; 2-water tank; 3-electromagnetic flowmeter; 4-first electromagnetic valve; 5-intermediate pipe; 6-nozzle; 7-second electromagnetic valve; 8-second air intake pipe; 9-air intake port; 10-third electromagnetic valve; 11-temperature and humidity flowmeter; 13-annular flow channel; 14-Venturi tube; 141-contraction tube; 142-throat; 143-diffuser; 15-nozzle pipe; 16-pneumatic material conveying pipe. DETAILED DESCRIPTION

[0052] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0054] Example 1

[0055] like Figure 1 , a schematic diagram of the structure of a water film generating device for a pneumatic material conveying system, comprising a first air intake pipe 1, a water tank 2, an electromagnetic flowmeter 3, a first electromagnetic valve 4, a nozzle pipe 5, a nozzle 6, a second electromagnetic valve 7, a second air intake pipe 8, an air intake port 9, a third electromagnetic valve 10, and a temperature and humidity flowmeter 11. The air intake port 9 is connected to the third electromagnetic valve 10 through a pipe, and the third electromagnetic valve 10 controls the compressed gas to open or close. The third electromagnetic valve 10 is connected to the first air intake pipe through a pipe. The first air intake pipe is used to balance the pressure of water in the water tank. The first air intake pipe is connected to the water tank 2 through a pipe. The water tank 2 is used to store water. The electromagnetic flowmeter 3 is connected to the water tank 2 through a flange. The electromagnetic flowmeter 3 is used to record the size of the water flow. The electromagnetic flowmeter 3 is connected to the first electromagnetic valve 4 through a flange. The first electromagnetic valve 4 is used to control the water injection rate and the amount of water injection. The first electromagnetic valve 4 is connected to the nozzle pipe 5 through a flange, and the nozzle pipe 5 is connected to the nozzle 6 through a three-way pipe.

[0056] like Figure 2, the nozzle structure includes an annular flow channel 13 and a venturi tube 14. The venturi tube 14 consists of three parts: a contraction tube 141, a throat 142 and a diffuser 143. Utilizing its structural characteristics, when the water-gas mixture flows through the contraction tube 141, the cross-sectional area of ​​the pipeline gradually decreases, the flow rate gradually increases, and the pressure decreases. When reaching the throat 142, since the throat 142 is the part with the smallest diameter in the structure of the venturi tube 14, the flow rate of the fluid reaches the maximum and the pressure reaches the minimum. The main function of the throat 142 is to maintain the high-speed flow state of the fluid and generate a large pressure difference therein. Finally, there is the diffuser 143, whose function is to reduce the flow rate, increase the pressure, and fully atomize the water flow. In addition, the number of Venturi tubes 14 arranged on the annular pipe surface is 16 nozzles per circle, and each Venturi tube 14 forms an angle of 45° with the horizontal plane. This is to make full use of the space to adjust the uniformity of water injection and ensure the water injection effect. At the same time, when the water-containing air flow is injected into the conveying pipe, it can rely on inertia force to better form a layer of water film on the pipe wall.

[0057] like Figure 3 , an intelligent control method for air moisture content in pneumatic material conveying pipelines, uses a recurrent neural network algorithm to build a prediction model for air saturation moisture content and pressure loss during pneumatic material conveying. The algorithm implementation process is as follows: a recurrent neural network is a kind of x- (1) ,…,x (t) A neural network model developed for sequence data. The recurrent neural network needs to target x (1) ,…,x (t) Sequence data has a certain logical association, that is, the state at time t that needs to be obtained only depends on the past x (1) ,…,x (t－1) and the input x at time t t The predicted value y(t) at time t depends on the previous x. (1) ,…,x (t－1) The corresponding y (1) ,…,y (t－1) and the input x at time t t Therefore, the present invention chooses to use a three-layer perceptual recurrent neural network to construct a prediction model for air saturated moisture content and pressure loss during pneumatic conveying, which includes an input layer, a hidden layer and an output layer. Neurons in the same layer are not connected to each other, while neurons between two adjacent layers are fully connected to each other. Pressure P, temperature T, humidity AH and air velocity v are used as input layers x1, x2, x3, x4, the predicted target air saturated moisture content is used as output layer y1, and the predicted pressure loss is used as output layer y2. After determining the input layer and output layer, in the forward propagation stage, the input layer obtains the input signal, that is, all influencing factors x1,…,x1 that are interrelated with air saturated moisture content and pressure loss. n, and pass it to each neuron in the hidden layer. Then, the hidden layer processes these signals and passes the processing results to the output layer. The back propagation algorithm (BP algorithm) is used to iterate and fit the data, that is, to solve the local extreme value by gradient descent, for each influencing factor x1, ..., x n , each neuron in the hidden layer has a corresponding input and output signal, marked as u j and h j , these two signals can be calculated by the following formulas:

[0058]

[0059] w ij is the weight between input layer neuron i and hidden layer neuron j; θ j is the random disturbance term,

[0060] The input signal lk and output signal Ck of the output layer can be calculated by the following formulas:

[0061]

[0062] Where: V jk is the weight between hidden layer neuron j and output layer neuron k; γ k is a random disturbance term; the output signal C of the output layer k The required prediction target is the air saturated moisture content; the input signal l k is the output signal h corresponding to each neuron in the hidden layer j The linear combination of the forward propagation output signal C k The difference from the true signal is the deviation. The deviation of each output layer neuron dk and hidden layer neuron deviation can be calculated by the following formula:

[0063] d k =(y k -C k )C k (1-C k ),k=1 (6)

[0064]

[0065] The bias is fed back from the output layer to the hidden layer, and the connection weights between the output layer and each hidden neuron are updated using the following formula (8). The connection weights between each hidden layer neuron and each influencing factor x1, ..., x2 in the input layer are updated using the following formula (8). n The weight between .

[0066]

[0067] Where: i = 1, 2, ..., n; j = 1, 2, 3, 4; k = 1; a1 and a2 are learning rates ranging from 0 to 1. The trained model is deployed in the actual pneumatic conveying system for real-time monitoring, prediction and optimization control.

[0068] like Figure 4 , an intelligent control method for the water content of air in a pneumatic material conveying pipeline, using a PID controller to adjust the first electromagnetic valve to adjust the water injection flow in real time. The PID controller is a very classic controller, which consists of three parts: proportional (P), integral (I) and differential (D). It can control the system deviation, cumulative deviation and deviation change rate respectively. Its expression is:

[0069]

[0070] Where u(t) is the control signal, e(t) is the error signal, and K p , K i , K d is a parameter of the PID controller. If the monitored pressure loss is lower than the pressure loss corresponding to the last system parameter setting value, the water injection amount is adjusted positively, otherwise it is reduced. By adjusting the response control signal, the water injection amount of the pneumatic conveying process is dynamically monitored and adjusted, thereby improving the operating efficiency and stability of the system. The intelligent control method of the water film generating device is centered on the monitoring device of the pneumatic material conveying system, and the specific steps are as follows:

[0071] (1) Initially set the air flow rate and particle feeding rate;

[0072] (2) Collect the pressure, temperature, humidity and air velocity before and after the material shoe and before and after the transportation;

[0073] (3) Calculate the saturated water content of air;

[0074] (4) Set the water injection flow rate based on the calculated air saturated water content.

[0075] (5) performing high-pressure spraying based on the set water injection flow rate;

[0076] (6) Determine whether the system pressure loss has decreased?

[0077] (7) If not, use PID controller to fine-tune and reduce the water injection amount;

[0078] (8) If yes, use PID controller to fine-tune and increase the water injection amount;

[0079] (9) Record the adjusted water injection volume, pressure, temperature, humidity and gas velocity.

[0080] (10) Inputting the recurrent neural network for training, predicting the output air saturation moisture content and pressure loss, and correcting the initial air flow rate and particle descent rate;

[0082] (11) Determine whether the delivery is completed?

[0083] (12) If not, return to step (2);

[0084] (13) If yes, flush the pipes and nozzles of the water film generating device;

[0085] (12)End.

[0086] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. A water film generating device for a pneumatic material conveying system, characterized in that: The invention comprises an air intake port (9), a third electromagnetic valve (10), a first air intake pipeline (1), a water tank (2), a first electromagnetic valve (4), an intermediate pipeline (5), a nozzle pipeline (15) and a nozzle (6); the input end of the first air intake pipeline (1) is connected to the air intake port (9) for inputting compressed gas, the third electromagnetic valve (10) is arranged on the first air intake pipeline (1), the output end of the first air intake pipeline (1) is inserted into the input port of the water tank (2) for balancing the water pressure in the water tank, the output port of the water tank (2) is connected to the input end of the intermediate pipeline (5) through the first electromagnetic valve (4), the first electromagnetic valve (4) is used to adjust the water injection flow, the output end of the intermediate pipeline (5) is connected to the nozzle (6) through the nozzle pipeline (15), and the nozzle (6) is used to provide a water film for the pneumatic material conveying pipeline (16).

2. The water film generating device according to claim 1, characterized in that: It also includes a second air intake pipeline (8), on which a second electromagnetic valve (7) is arranged, the input end of the second air intake pipeline (8) is connected to the first air intake pipeline (1) near the air intake port (9), and the output end of the second air intake pipeline (8) is connected to the intermediate pipeline (5).

3. The water film generating device according to claim 2, characterized in that: When the second electromagnetic valve (7) is opened alone, it is used to flush the pipeline and nozzle of the water film generating device.

4. The water film generating device according to claim 1, characterized in that: It also includes an electromagnetic flowmeter (3) for recording the size of the water flow, and the electromagnetic flowmeter (3) is connected to the water tank (2) and the first electromagnetic valve (4) respectively through flanges.

5. The water film generating device according to claim 1, characterized in that: The nozzle pipe (15) is at least two-way pipes, the nozzle (6) is at least two nozzles, the nozzle pipes are respectively connected to the corresponding nozzles (6), the nozzle (6) is an annular nozzle, which comprises an annular pipe (13) and a venturi tube (14) connected in sequence, and the venturi tube (14) is a plurality of venturi tubes (14) uniformly arranged on the annular pipe (13).

6. The water film generating device according to claim 5, characterized in that: Each of the venturi tubes (14) is arranged to form an angle of 45° with the horizontal plane. The venturi tube (14) comprises a contraction tube (141), a throat tube (142) and a diffusion tube (143) which are connected in sequence. Among the three, the throat tube (142) has the smallest diameter.

7. A method for intelligently controlling the moisture content of air in a pneumatic material conveying pipeline, the method being applied to the water film generating device described in any one of claims 1 to 6, using a PID controller to control the first electromagnetic valve (4) to adjust the water injection flow in real time, using a recurrent neural network to construct a prediction model for the saturated moisture content and pressure loss of air during pneumatic material conveying, inputting the recorded pressure P, temperature T, humidity AH and air velocity V into the prediction model, and correcting the initial air flow and particle feeding rate based on the prediction results of the recurrent neural network until the material conveying is completed and the control is terminated.

8. According to the intelligent control method of the air moisture content in the pneumatic material conveying pipeline according to claim 7, the recurrent neural network includes an input layer, a hidden layer and an output layer, based on the pressure P, temperature T, humidity AH and air velocity V of the pneumatic material conveying pipeline as input layer parameters, the forward propagation algorithm and the backward propagation algorithm are used in the network model to perform data iteration and fitting, and the gradient descent is used to solve the local extreme value to complete the training of the model and predict the output air saturated moisture content and pressure loss.

9. According to the intelligent control method of air moisture content in a pneumatic material conveying pipeline as claimed in claim 7, before using the PID controller to control the first electromagnetic valve (4) to adjust the injection flow in real time, the following steps are further performed: (1) Initially set the air flow rate and particle feeding rate; (2) Collect the pressure, temperature, humidity and air velocity before and after the material shoe and before and after the transportation; (3) Calculate the saturated water content of air; (4) Set the water injection flow rate based on the calculated air saturated water content.

10. The intelligent control method for air moisture content in a pneumatic material conveying pipeline according to claim 9, wherein the use of a PID controller to control the first electromagnetic valve (4) to adjust the injection flow in real time specifically comprises the following steps: (1) High-pressure spraying based on the set water injection flow rate; (2) Determine whether the system pressure loss has decreased; (3) If not, use PID controller to fine-tune and reduce the water injection flow rate; (4) If yes, use PID controller to fine-tune and increase the water injection flow rate; (5) Record the adjusted water injection volume, pressure, temperature, humidity and gas velocity.

Images