Digital energy air compression station water circulation system capable of preventing blockage

Abstract

The present application relates to the field of air compression station, especially to a digital energy air compression station water circulation system capable of preventing blockage, which is used for water circulation system of air compression station heat dissipation, comprising: intelligent regulating valve, water quality softening tank, multifunctional intelligent control system, wherein: the intelligent regulating valve, the multifunctional water circulation system and the water quality softening tank pipeline are connected, the intelligent regulating valve is wirelessly connected with the multifunctional intelligent control system; the multifunctional intelligent control system comprises a data collection module, an evaluation module and an execution module, the evaluation module analyzes and judges each sensing value based on a technical solution improved based on a neural network algorithm, and the execution module is used for executing the judgment of the evaluation module to control the water quality softening tank for easy descaling; the present application obtains optimal control measures through the evaluation module to efficiently soften water quality and avoid blockage, and through the self-checking module, error points are easily obtained and timely notified, so as to achieve efficient operation of the air compression station.

Description

Technical Field

[0001] This invention relates to the field of air compressor stations, and more particularly to a digital energy air compressor station water circulation system that can achieve anti-clogging. Background Technology

[0002] Compressed air, as the most environmentally friendly power source in industrial production, is widely used in various industries such as pharmaceuticals, food, machinery, electronics, plastics, textiles, power, and building materials. However, during actual operation, air compressors generate a large amount of heat. In order for the air compressors to operate efficiently for a long time, a water circulation system is often used to cool them down. However, due to the influence of water quality in the water circulation system, long-term operation of the air compressor station often leads to the formation of scale, which blocks the water circulation system and indirectly affects the service life of the air compressor. Moreover, during maintenance, the water circulation system of the air compressor station needs to be shut down, disassembled, and the blockage points need to be checked and cleared one by one, which is costly and seriously affects the factory's production efficiency. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides a digital energy air compressor station water circulation system that can prevent clogging, eliminate scale in the water circulation system, and perform self-checks on the water circulation system during air compressor station operation.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] A clog-resistant digital energy air compressor station water circulation system, used for heat dissipation in the air compressor station, includes: an intelligent regulating valve, a water softening tank, and a multi-functional intelligent control system, wherein: the intelligent regulating valve, the multi-functional water circulation system, and the water softening tank are interconnected, and the intelligent regulating valve is wirelessly connected to the multi-functional intelligent control system; wherein:

[0006] The multi-functional intelligent control system includes a data collection module, an evaluation module, and an execution module;

[0007] The data collection module is used to collect temperature, pressure, water inflow, gas consumption, gas storage, magnesium ion and calcium ion values ​​from each sensor in real time. Each sensor is named according to its location.

[0008] The evaluation module is used for preprocessing and analyzing the data from each sensor to select the optimal control measures.

[0009] The execution module is used to control the judgment of the evaluation module. When the evaluation module determines that the scaling conditions are met or the water quality is too hard and the corresponding value meets the softening condition, the execution module is activated to control the intelligent regulating valve and open the water softening tank to soften the water circulation system. When the evaluation module determines that the water quality does not need to be adjusted, the execution module is activated to control the intelligent regulating valve and close the water softening tank.

[0010] Specifically, the evaluation module's specific methods include the following steps:

[0011] S1. Cluster analysis is used to classify magnesium ion and calcium ion data to obtain ion indices; the ratio of water temperature to water pressure is calculated to obtain water environment indices.

[0012] S2. Using scale formation in the water cycle as the explanatory variable and ion index and water environment index as the response variables, a random forest model is constructed to rank the importance of scale formation in the water cycle in order to identify key scaling information.

[0013] S3. Using key scaling information as input variables and ion index and water environment index as output variables, a BP artificial neural network model is constructed to quantify the response relationship between the two.

[0014] Further, step S1 includes the following steps:

[0015] S11. Normalize the magnesium ion value and the calcium ion value.

[0016] S12. Randomly select several data objects in the data space as initial cluster centers;

[0017] S13. Calculate the Euclidean distance between all data objects and the initial cluster center, and assign each data object to the initial cluster center with the smallest Euclidean distance to form the initial cluster.

[0018] S14. Calculate the average value of the data objects in each initial cluster and use the average value as the new cluster center;

[0019] S15. Repeat the calculation of the Euclidean distance between all data objects and the new cluster center, and assign each data object to the new cluster center with the smallest Euclidean distance to form a new cluster. Calculate the average value of each new cluster as the new cluster center again until the new cluster centers no longer change, determine the final cluster, and determine the ion index corresponding to each final cluster based on the magnesium and calcium ion data characteristics of the final cluster.

[0020] Further, step S2 includes the following steps:

[0021] S21. Randomly select L samples from the dataset to serve as the training set;

[0022] S22. Randomly and without repetition, select N variables related to scale formation in the water cycle as split points, and use the Gini coefficient to determine the optimal split points for the ion index and water environment index variables to generate a decision tree.

[0023] S23. Repeat steps S21-S22 M times to obtain M decision trees and generate a random forest.

[0024] S24. Calculate the prediction error of each decision tree in the random forest for the out-of-bag data;

[0025] S25. Add random perturbation to the variable of scale formation in water circulation in the data outside the bag, and calculate the prediction error of each decision tree in the random forest for the random perturbation.

[0026] S26. Rank the importance of the scale formation variables in the water cycle according to the prediction error of the random forest for the data outside the bag;

[0027] S27. Select the top n water circulation scale formation data based on importance as key scale formation information.

[0028] Furthermore, step S3 includes the following steps:

[0029] S31. Determine the number and quantity of neurons in the neural network. The neurons include an input layer, a hidden layer, and an output layer, wherein the input layer contains key scaling information, and the output layer contains ion index and water environment index.

[0030] S32. The input data is processed through a neural network, starting from the input layer and passing through the hidden layers until it reaches the output layer. In each neuron, a weighted sum is calculated based on the input data and weights, and finally the output of the neuron is obtained through an activation function.

[0031] S33. Compare the output of the neural network with the expected output and calculate the error;

[0032] It should be noted that the expected output here refers to the actual value of the output vector sample;

[0033] S34. Based on the error, starting from the output layer, propagate the error backward layer by layer, and adjust the connection weights and biases using the gradient descent algorithm based on the error.

[0034] S35. Repeat steps S32 to S34 until the error reaches the preset error threshold.

[0035] S36. Test the trained neural network using an independent test dataset and evaluate the model's predictive performance using the coefficient of determination.

[0036] Furthermore, it also includes a filtration system, which comprises an impurity filtration device and an ultraviolet sterilization device; the impurity filtration device is used to filter solid particles in the water circulation system to prevent them from accumulating and forming blockages, and the ultraviolet sterilization device is used to kill microorganisms in the water to prevent them from multiplying and causing blockages.

[0037] Furthermore, the softening solution in the water softening tank is one of the following: sodium chloride solution, calcium hydroxide preparation, sodium carbonate preparation, or diluted sodium tripolyate solution.

[0038] Furthermore, it also includes a heat consumption system, which is used to heat the plant equipment through the thermal effect of the water circulation system. When the heat consumption system needs to start heating, it controls the intelligent regulating valve to connect the water circulation system to the heat consumption system, and after completing the heat supply, it flows back to the water circulation system.

[0039] Furthermore, it also includes air compressors, air tanks, and sensing systems, based on which the maximum and minimum historical air consumption values ​​are obtained;

[0040] When the air tank reaches its maximum storage capacity, the air compressor stops working; when the air tank reaches its minimum storage capacity, the air compressor starts working.

[0041] Furthermore, the multi-functional intelligent control system also includes a water volume control system, which is connected to the water inlet. The water volume control system obtains the water volume value through the data collection module and compares it with the pre-set water circulation system water storage threshold and the minimum water storage value.

[0042] When the inlet water volume is close to or exceeds the water storage threshold of the water circulation system, the intelligent regulating valve is reduced until it is closed; when the water storage volume is less than the minimum water storage value, the intelligent regulating valve opens the inlet pipe to supply water.

[0043] Furthermore, the multi-functional intelligent control system also includes a self-test module and an alarm module;

[0044] The self-test module is used to check for faults in the entire water circulation system equipment using water pressure, specifically:

[0045] The first and second water pressure thresholds are set based on historical values.

[0046] When the water pressure reading at a certain point exceeds the first water pressure threshold, it is determined that the nearby water pipes are blocked.

[0047] When the water pressure value at a certain point is lower than the second water pressure threshold, it is determined that the water pipe near the sensor is leaking or the front end of that section of the pipe is blocked.

[0048] When all water pressure values ​​are higher than the first water pressure threshold, it is determined that the intelligent regulating valve or water volume control system is malfunctioning.

[0049] When all water pressure values ​​are lower than the second water pressure threshold, it can be determined that the intelligent regulating valve is faulty or the inlet water volume is set incorrectly.

[0050] The alarm module is used to notify management personnel to adjust and repair the faulty location when the self-test module determines that the water pressure value is higher than the first water pressure threshold or lower than the second water pressure threshold.

[0051] The beneficial effects of this invention are as follows:

[0052] 1. By using a technical solution based on improved neural network algorithms, we can efficiently determine and quantify the impact of the interaction between calcium ions, magnesium ions, water temperature, and water pressure on scaling, and then regulate the water softening tank to soften the water in the water circulation system, thereby preventing scale formation and reducing blockages.

[0053] 2. By utilizing the heat consumption system to reuse the hot water generated in the air compressor station and by rationally controlling the start-up and shutdown of the air compressor and water storage, efficiency can be improved while reducing costs and unnecessary waste.

[0054] 3. Self-inspection using water pressure values ​​efficiently screens faulty facilities, avoiding downtime caused by troubleshooting one by one, thus reducing the impact on production. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the multifunctional command and control system module of the present invention;

[0056] Figure 2 This is a structural diagram of the air compressor water-air pipeline system of the present invention;

[0057] Figure 3 This is a flowchart of the evaluation module of the present invention.

[0058] In the diagram: 1. Air compressor; 2. Air tank; 3. Dryer; 4. Cooling system; 5. Filtration system; 6. Sensing system; 7. Water control system; 8. Intelligent regulating valve; 9. Heat consumption system; 10. Water softening tank. Detailed Implementation

[0059] Please see Figure 1 , Figure 2 and Figure 3 As shown, this invention provides a clog-resistant digital energy air compressor station water circulation system, including an air compressor 1, an air storage tank 2, a dryer 3, a cooling system 4, a filtration system 5, a sensing system 6, a water flow control system 7, an intelligent regulating valve 8, a heat consumption system 9, a water softening tank 10, a piping system, and a multi-functional intelligent control system. The air compressor 1, air storage tank 2, dryer 3, cooling system 4, filtration system 5, sensing system 6, water flow control system 7, intelligent regulating valve 8, multi-functional water circulation system, and water softening tank 10 are connected in a closed loop via the piping system. The sensing system 6 and intelligent regulating valve 8 are wirelessly connected to the multi-functional intelligent control system.

[0060] The multi-functional intelligent control system includes a data collection module, an evaluation module, and an execution module, as well as a storage module;

[0061] The data collection module is used to collect temperature, pressure, water inflow, gas consumption, gas storage, magnesium ion and calcium ion values ​​fed back by each sensor in the sensing system 6 in real time. Each sensor is named according to its location.

[0062] The storage module is used to store the data collected by the data collection module, the data of each mode set by the execution module, and to generate work logs;

[0063] The evaluation module is used to preprocess the collected data and perform algorithmic evaluation in order to select the optimal control measures;

[0064] The preprocessing includes: data cleaning, normalization, data transformation, feature engineering, and data discretization;

[0065] The execution module is used to control the judgment of the evaluation module. When the evaluation module determines that the scaling conditions are met or the water quality is too hard and the corresponding value meets the softening conditions, the execution module is activated to control the intelligent regulating valve 8 and open the water softening tank 10 to soften the water circulation system. When the evaluation module determines that the water quality does not need to be adjusted, the execution module is activated to control the intelligent regulating valve 8 and close the water softening tank 10.

[0066] Specifically, the evaluation module's specific methods include the following steps:

[0067] S1. Data Processing: Cluster analysis is used to classify magnesium ion and calcium ion data to obtain ion indices; the ratio of water temperature to water pressure is calculated to obtain the water environment index.

[0068] S2. Key Information Identification: Using scale formation in the water cycle as the explanatory variable and ion index and water environment index as the response variables, a random forest model is constructed to rank the importance of scale formation in the water cycle in order to identify key scale information.

[0069] S3. Key Relationship Model: Using key scaling information as input variables and ion index and water environment index as output variables, a BP artificial neural network model is constructed to quantify the response relationship between the two.

[0070] In step S1, the clustering analysis method is configured as the K-means clustering method, which includes the following steps:

[0071] S11. Normalization: Normalize the magnesium ion value and the calcium ion value.

[0072] S12. Initialize cluster centers: Randomly select several data objects in the data space as initial cluster centers;

[0073] S13. Initialize clusters: Calculate the Euclidean distance between all data objects and the initial cluster centers, and assign each data object to the initial cluster center with the smallest Euclidean distance to form the initial cluster.

[0074] S14. Update cluster centers: Calculate the average value of the data objects in each initial cluster and use this average value as the new cluster centers;

[0075] S15. Determine the final clusters: Repeat the calculation of the Euclidean distance between all data objects and the new cluster centers, and assign each data object to the new cluster center with the smallest Euclidean distance to form a new cluster; calculate the average value of each new cluster as the new cluster center again until the new cluster centers no longer change, determine the final clusters, and determine the ion index corresponding to each final cluster based on the magnesium ion and calcium ion data characteristics of the final clusters.

[0076] Further, in step S11, the normalization process is calculated using the following formula:

[0077]

[0078] In the formula, Y i This represents the normalized data sample of the i-th magnesium and calcium ions; K i For the i-th magnesium and calcium ion data sample; K max K represents the maximum value of all magnesium and calcium ion data samples; min This represents the minimum value among all magnesium and calcium ion data samples.

[0079] Further, in step S13, the Euclidean distance is calculated using the following formula:

[0080]

[0081] In the formula, d(x,C) i ) represents the Euclidean distance function; x is the normalized data object; C i Let x represent the i-th cluster center; n is the number of samples in the dataset; x j Ci represents the data object of the j-th variable in the dataset. j This represents the i-th cluster center of the j-th variable.

[0082] Further, in step S1, the water environment index is calculated using the following formula:

[0083] R k =L k / S k

[0084] In the formula, R k L represents the k-th water environment index;k S represents the k-th water temperature value; k This represents the k-th water pressure value.

[0085] In step S2, the construction of the random forest model to rank the importance of scale formation in the water cycle includes the following steps:

[0086] S21. Randomly select L samples from the dataset to serve as the training set;

[0087] S22. Randomly and without repetition, select N variables related to scale formation in the water cycle as split points, and use the Gini coefficient to determine the optimal split points for the ion index and water environment index variables to generate a decision tree.

[0088] S23. Repeat steps S21-S22 M times to obtain M decision trees and generate a random forest.

[0089] S24. Calculate the prediction error of each decision tree in the random forest for the out-of-bag data;

[0090] S25. Add random perturbation to the variable of scale formation in water circulation in the data outside the bag, and calculate the prediction error of each decision tree in the random forest for the random perturbation.

[0091] S26. Rank the importance of the scale formation variables in the water cycle according to the prediction error of the random forest for the data outside the bag;

[0092] S27. Select the top n water circulation scale formation data based on importance as key scale formation information.

[0093] It should be noted that for the top n water cycle scale formation variables in the importance ranking, n can depend on the value at which the prediction error begins to increase after the nth water cycle scale formation variable is added.

[0094] Furthermore, in step S23, the determination of the value of M for the M decision trees is as follows:

[0095] Using O decision trees as the step size, generate random forests with different numbers of decision trees;

[0096] Calculate the prediction error of the random forest for the test samples;

[0097] Select the random forest with the smallest prediction error and determine the number of decision trees M.

[0098] Furthermore, in step S26, the importance of scale formation in the water cycle is ranked based on the prediction error of the random forest on the data outside the bag. The calculation formula is as follows:

[0099]

[0100] In the formula: PIM represents the importance of the scale formation variable in the water cycle; M is the number of decision trees in the random forest; ErrOOB represents the prediction error of the k-th decision tree for out-of-bag data with random perturbation of the i-th water cycle scale formation variable; k Let represent the prediction error of the k-th decision on the out-of-bag data without added random perturbation.

[0101] In step S3, the BP artificial neural network model includes the following construction steps:

[0102] S31. Determine the model structure: Determine the number and quantity of neurons in the neural network. The neurons include an input layer, a hidden layer, and an output layer. The input layer contains key scaling information, and the output layer contains ion index and water environment index.

[0103] S32. Forward Propagation: The input data is processed through a neural network, starting from the input layer and passing through the hidden layers until it reaches the output layer. In each neuron, a weighted sum is calculated based on the input data and weights, and finally the output of the neuron is obtained through an activation function.

[0104] S33. Error Calculation: Compare the output of the neural network with the expected output and calculate the error.

[0105] It is important to note that the expected output here refers to the actual value of the output vector sample.

[0106] S34. Backpropagation: Based on the error, starting from the output layer, backpropagate the error layer by layer, and adjust the connection weights and biases using the gradient descent algorithm based on the error.

[0107] S35. Repeated iteration: Repeat steps S32 to S34 until the error reaches the preset error threshold.

[0108] S36. Model Evaluation: Test the trained neural network using an independent test dataset and evaluate the model's predictive performance using the coefficient of determination.

[0109] Further, in step S31, the model structure has five hidden layers, and the number of input and output layers is the same as the number of samples. Connection weights and biases are provided between input layer units and hidden layer units, and between hidden layer units and output layer units. Both hidden and output layers have activation functions. The calculation formulas for the neuron outputs of the hidden and output layers are as follows:

[0110]

[0111] In the formula, y is the neuron output; xi is the input; b is the bias; wi is the connection weight between neurons; F is the activation function; and k is the number of inputs.

[0112] Further, in step S33, the calculation error is calculated using the following formula:

[0113]

[0114] In the formula, E is the error function of the output layer unit; tl is the desired output; and zl is the output of the output layer.

[0115] Using the methods described above, we can analyze key scaling information that is highly efficient in understanding the impact of factors such as calcium ions, magnesium ions, water temperature, and water pressure on scale formation.

[0116] In the water circulation pipeline system of the present invention, water flows into the water circulation pipeline system through the water flow control system 7, and then enters the air compressor 1 through the sensing system 6 and the filtration system 5 for cooling. After flowing out of the air compressor 1, it passes through the sensing system 6 and the filtration system 5 again to cool the air storage tank 2. After flowing out, it passes through the sensing system 6 and the filtration system 5 again to the dryer 3 for cooling, and then flows to the cooling tower in the cooling system 4 to cool the water source. The cooled water source is then circulated again.

[0117] When the multi-functional control system determines that water softening is required, it opens the intelligent regulating valve 8 of the water softening tank 10, allowing the softening liquid in the water softening tank 10 to flow out and merge into the water circulation system for water softening. The softening liquid can be sodium chloride solution, calcium hydroxide preparation, or sodium carbonate preparation, preferably diluted sodium tripolyate solution, which can remove calcium and magnesium ions from the water and has the characteristics of softening water, dispersing dirt, acting as an alkali buffer, and preventing caking. When the evaluation module determines that the water quality does not need to be adjusted, it activates the execution module to regulate the intelligent regulating valve and close the water softening tank. The water softening tank 10 is normally in the closed state.

[0118] When the heat consumption system 9 needs heat supply, the intelligent regulating valve 8 near the heat consumption system 9 needs to be opened and the water circulation system of the air compressor 1 to the air storage tank 2 needs to be closed. The water circulation system changed in this way can provide heat to the heat consumption system 9. After the water flowing through the heat consumption system 9 completes the heat supply, it flows back to the air storage tank 2 through the sensing system 6 and the filtration system 5 to continue the cooling work.

[0119] The heat consumption system 9 includes a heating unit and a thermoelectric conversion unit, which reuses the hot water generated by the air compressor station to improve economic efficiency.

[0120] It should be noted that the heat dissipation system 9 cannot be a shower unit, because when the water softening tank 10 is turned on for descaling, the softening liquid inside the water softening tank 10 will have an impact on human health.

[0121] Based on the obtained water inflow value, compare it with the set water storage value of the water circulation system, and control the water inflow value to be less than or equal to the water storage value of the water circulation system; the multi-functional intelligent control system also includes a water control system, which is connected to the water inlet. The water inflow value obtained through the data collection module is compared with the pre-set water storage threshold and minimum water storage value of the water circulation system.

[0122] When the inlet water volume is close to or exceeds the water storage threshold of the water circulation system, the intelligent regulating valve is reduced until it is closed; when the water storage volume is less than the minimum water storage value, the intelligent regulating valve opens the inlet pipe to supply water.

[0123] Water pressure sensors are installed at various equipment and water pipes in the water circulation pipeline system to monitor the water circulation status.

[0124] The multi-functional intelligent control system also includes a self-test module and an alarm module;

[0125] The self-test module is used to check for faults in the entire water circulation system equipment using water pressure, specifically:

[0126] The first and second water pressure thresholds are set based on historical values.

[0127] When the water pressure reading at a certain point exceeds the first water pressure threshold, it is determined that the nearby water pipes are blocked.

[0128] When the water pressure value at a certain point is lower than the second water pressure threshold, it is determined that the water pipe near the sensor is leaking or the front end of that section of the pipe is blocked.

[0129] When all water pressure values ​​are higher than the first water pressure threshold, it is determined that the intelligent regulating valve or water volume control system is malfunctioning.

[0130] When all water pressure values ​​are lower than the second water pressure threshold, it can be determined that the intelligent regulating valve is faulty or the inlet water volume is set incorrectly.

[0131] The alarm module is used to notify management personnel to adjust and repair the faulty location when the self-test module determines that the water pressure value is higher than the first water pressure threshold or lower than the second water pressure threshold.

[0132] Among them, the first water pressure threshold can be set to 0.55 MPa and the second water pressure threshold to 0.4 MPa based on the conventional water pressure statistics of the industrial water circulation system. However, the water pressure threshold of each industrial water circulation system will fluctuate to some extent depending on factors such as the height of the industrial building, the performance of the water supply equipment and the fluctuation of water demand.

[0133] The present invention provides intelligent regulating valves 8, sensing systems 6 and filtration systems 5 in various devices and water circulation pipelines in the water circulation pipeline system;

[0134] The intelligent regulating valve 8 includes: intelligent transfer valve and intelligent on / off valve. All of these valves can be remotely controlled via transfer switch.

[0135] The sensing system 6 includes: a smart pressure sensor, a smart temperature sensor, a smart pH sensor, a smart calcium ion sensor, and a smart magnesium ion sensor. All sensors involved in the sensing system 6 can remotely transmit sensing values.

[0136] The filtration system 5 includes impurity filtration equipment and ultraviolet sterilization equipment. Impurity filtration equipment is used to filter solid particles in the water circulation system to prevent them from accumulating and forming blockages. Common equipment includes bag filters, cartridge filters, etc. Ultraviolet sterilization equipment is used to kill microorganisms in the water to prevent them from multiplying and causing blockages. Equipment such as ultraviolet lamps is used to kill microorganisms in the water.

[0137] Cooling system 4 includes: cooling tower and water circulation system.

[0138] In the gas pipeline system of this invention, air is compressed by air compressor 1 and enters the first air storage tank 2. Oil and gas separation is carried out in the air storage tank 2. The separated oil is discharged from the bottom of the air storage tank 2. The separated compressed air enters the dryer 3 for drying and dehydration. The dried compressed air finally enters the second air storage tank 2 for later use.

[0139] To meet different air usage requirements, appropriate gas filtration equipment can be added to the gas pipeline system between air compressor 1 and the second air tank 2 to ensure the quality requirements of compressed air.

[0140] The air compressor 1, air tank 2, dryer 3, and filtration system 5 are all equipped with a sensing system 6 and an intelligent regulating valve 8.

[0141] In the gas pipeline system, sensing system 6 needs to be equipped with an intelligent humidity sensor, and the data should be transmitted to the multi-functional intelligent control system via wireless communication.

[0142] The intelligent regulating valves 8 are all installed at the inlet and outlet of the gas passages of each device and on the pipelines of the gas pipeline system.

[0143] Furthermore, both the sensing system 6 and the intelligent regulating valve 8 can be remotely controlled to acquire and adjust data, preferably using NB-IoT transmission.

[0144] The sensor values ​​for gas consumption and storage are transmitted to the multi-functional control system for comparison and analysis. The maximum and minimum historical gas consumption values ​​can be determined. When the gas storage tank 2 reaches the maximum gas consumption value, the air compressor 1 can be stopped. When the gas storage tank 2 reaches the minimum gas consumption value, the air compressor 1 can be started. The above method helps to save industrial electricity.

[0145] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A water circulation system for a digital energy air compressor station that can achieve anti-clogging, characterized in that, The digital energy air compressor station water circulation system, used for heat dissipation in the air compressor station, includes: an intelligent regulating valve, a water softening tank, and a multi-functional intelligent control system. The intelligent regulating valve, the multi-functional water circulation system, and the water softening tank are interconnected, and the intelligent regulating valve is wirelessly connected to the multi-functional intelligent control system. The multi-functional intelligent control system includes a data collection module, an evaluation module, and an execution module; The data collection module is used to collect temperature, pressure, water inflow, gas consumption, gas storage, magnesium ion and calcium ion values ​​from each sensor in real time. Each sensor is named according to its location. The evaluation module is used for preprocessing and analyzing the data from each sensor to select the optimal control measures. The execution module is used to control the judgment of the evaluation module. When the evaluation module determines that the scaling conditions are met or the water quality is too hard and the corresponding value meets the softening condition, the execution module is activated to control the intelligent regulating valve and open the water softening tank to soften the water circulation system. When the evaluation module determines that the water quality does not need to be adjusted, the execution module is activated to control the intelligent regulating valve and close the water softening tank. Specifically, the evaluation module's specific methods include the following steps: S1. Cluster analysis is used to classify magnesium ion and calcium ion data to obtain ion indices; the ratio of water temperature to water pressure is calculated to obtain water environment indices. S2. Using scale formation in the water cycle as the explanatory variable and ion index and water environment index as the response variables, a random forest model is constructed to rank the importance of scale formation in the water cycle in order to identify key scaling information. S3. Using key scaling information as input variables and ion index and water environment index as output variables, a BP artificial neural network model is constructed to quantify the response relationship between the two. Step S1 includes the following steps: S11. Normalize the magnesium ion value and the calcium ion value. S12. Randomly select several data objects in the data space as initial cluster centers; S13. Calculate the Euclidean distance between all data objects and the initial cluster center, and assign each data object to the initial cluster center with the smallest Euclidean distance to form the initial cluster. S14. Calculate the average value of the data objects in each initial cluster and use the average value as the new cluster center; S15. Repeat the calculation of the Euclidean distance between all data objects and the new cluster center, and assign each data object to the new cluster center with the smallest Euclidean distance to form a new cluster. Calculate the average value of each new cluster as the new cluster center again until the new cluster centers no longer change, determine the final cluster, and determine the ion index corresponding to each final cluster based on the magnesium and calcium ion data characteristics of the final cluster.

2. The digital energy air compressor station water circulation system according to claim 1, characterized in that, Step S2 includes the following steps: S21. Randomly select L samples from the dataset to serve as the training set; S22. Randomly and without repetition, select N variables related to scale formation in the water cycle as split points, and use the Gini coefficient to determine the optimal split points for the ion index and water environment index variables to generate a decision tree. S23. Repeat steps S21-S22 M times to obtain M decision trees and generate a random forest. S24. Calculate the prediction error of each decision tree in the random forest for the out-of-bag data; S25. Add random perturbation to the variable of scale formation in water circulation in the data outside the bag, and calculate the prediction error of each decision tree in the random forest for the random perturbation. S26. Rank the importance of the scale formation variables in the water cycle according to the prediction error of the random forest for the data outside the bag; S27. Select the top n water circulation scale formation data based on importance as key scale formation information.

3. The digital energy air compressor station water circulation system according to claim 1, characterized in that, Step S3 includes the following steps: S31. Determine the number and quantity of neurons in the neural network. The neurons include an input layer, a hidden layer, and an output layer, wherein the input layer contains key scaling information, and the output layer contains ion index and water environment index. S32. The input data is processed through a neural network, starting from the input layer and passing through the hidden layers until it reaches the output layer. In each neuron, a weighted sum is calculated based on the input data and weights, and finally the output of the neuron is obtained through an activation function. S33. Compare the output of the neural network with the expected output and calculate the error; S34. Based on the error, starting from the output layer, propagate the error backward layer by layer, and adjust the connection weights and biases using the gradient descent algorithm based on the error. S35. Repeat steps S32 to S34 until the error reaches the preset error threshold. S36. Test the trained neural network using an independent test dataset and evaluate the model's predictive performance using the coefficient of determination.

4. The digital energy air compressor station water circulation system according to claim 1, characterized in that: It also includes a filtration system, which comprises an impurity filtration device and an ultraviolet sterilization device; the impurity filtration device is used to filter solid particles in the water circulation system to prevent them from accumulating and forming blockages, and the ultraviolet sterilization device is used to kill microorganisms in the water to prevent them from multiplying and causing blockages.

5. The digital energy air compressor station water circulation system according to claim 4, characterized in that, The softening solution in the water softening tank is one of the following: sodium chloride solution, calcium hydroxide preparation, sodium carbonate preparation, or diluted sodium tripolyate solution.

6. The digital energy air compressor station water circulation system according to claim 1, characterized in that, It also includes a heat consumption system, which is used to heat the plant equipment through the thermal effect of the water circulation system. When the heat consumption system needs to start heating, it controls the intelligent regulating valve to connect the water circulation system to the heat consumption system, and after completing the heat supply, it flows back to the water circulation system.

7. The digital energy air compressor station water circulation system according to claim 1, characterized in that, It also includes air compressors, air tanks, and sensing systems, which are used to determine the maximum and minimum historical air consumption. When the air tank reaches its maximum storage capacity, the air compressor stops working; when the air tank reaches its minimum storage capacity, the air compressor starts working.

8. The digital energy air compressor station water circulation system according to claim 1, characterized in that, The multi-functional intelligent control system also includes a water volume control system, which is connected to the water inlet. The water volume control system obtains the water volume value through the data collection module and compares it with the pre-set water circulation system water storage threshold and the minimum water storage value. When the inlet water volume is close to or exceeds the water storage threshold of the water circulation system, the intelligent regulating valve is reduced until it is closed; when the water storage volume is less than the minimum water storage value, the intelligent regulating valve opens the inlet pipe to supply water.

9. The digital energy air compressor station water circulation system according to claim 1, characterized in that, The multi-functional intelligent control system also includes a self-test module and an alarm module; The self-test module is used to check for faults in the entire water circulation system equipment using water pressure, specifically: The first and second water pressure thresholds are set based on historical values. When the water pressure reading at a certain point exceeds the first water pressure threshold, it is determined that the nearby water pipes are blocked. When the water pressure value at a certain point is lower than the second water pressure threshold, it is determined that the water pipe near the sensor is leaking or the front end of the pipe is blocked. When all water pressure values ​​are higher than the first water pressure threshold, it is determined that the intelligent regulating valve or water volume control system is malfunctioning. When all water pressure values ​​are lower than the second water pressure threshold, it is determined that the intelligent regulating valve is faulty or the inlet flow rate is set incorrectly. The alarm module is used to notify management personnel to adjust and repair the faulty location when the self-test module determines that the water pressure value is higher than the first water pressure threshold or lower than the second water pressure threshold.

Images