Energy-saving control method and system for compressed air system

Abstract

The invention provides an energy-saving control method and system for a compressed air system, and belongs to the technical field of compressed air systems. Comprising the following steps: S1, acquiring various parameters of a compressed air system, and constructing a digital twinborn body of the compressed air system; s2, acquiring real-time data and historical data of a compressed air system based on the digital twinborn body, establishing a gas consumption prediction model according to the real-time data and the historical data, and predicting a gas consumption demand in a future period based on the gas consumption prediction model; s3, on the premise of meeting the gas consumption demand in the future time period, taking the minimization of the total energy consumption as the target, establishing a target function, and setting the constraint condition of the target function; s4, solving the objective function to obtain an optimal control strategy, meeting constraint conditions, of the compressed air system; and S5, the compressed air system is controlled based on the optimal control strategy, so that the energy consumption of the compressed air system is effectively reduced while the compressed air system meets the air consumption requirement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of compressed air systems, in particular to a compressed air system energy-saving control method and system. BACKGROUND

[0002] The single crystal furnace opens or closes the main vacuum valve through the pneumatic control device, and the pneumatic control device is usually driven by compressed air. Continuous supply of compressed air can ensure normal production of the single crystal furnace. However, under the management mode of the traditional compressed air system, most enterprises adopt a rough control strategy, that is, the compressed air system works according to a fixed operation mode, and lacks accurate prediction and dynamic response mechanism for air demand, so that in the low load scene, the air compressor still maintains high load operation, and continuously produces far more compressed air than the actual demand. A large amount of compressed air is wasted in the pressure relief process, causing inefficient consumption of energy. SUMMARY

[0003] Therefore, the present application provides a compressed air system energy-saving control method and system to solve the technical problem of energy waste in the traditional compressed air system management mode.

[0004] The technical scheme adopted by the present application to solve its technical problems is:

[0005] A compressed air system energy-saving control method, comprising.

[0006] S1, acquiring parameters of a compressed air system, and constructing a digital twin of the compressed air system;

[0007] S2, obtaining real-time data and historical data of the compressed air system based on the digital twin, establishing an air consumption prediction model according to the real-time data and the historical data, and predicting air consumption demand in a future period based on the air consumption prediction model;

[0008] S3, under the premise of meeting the air consumption demand in the future period, minimizing the total energy consumption as the target, establishing a target function, and setting the constraint conditions of the target function;

[0009] S4, solving the target function to obtain an optimal control strategy of the compressed air system that meets the constraint conditions;

[0010] S5, controlling the compressed air system based on the optimal control strategy.

[0011] Preferably, in step S1, the parameters of the compressed air system are acquired, and the digital twin of the compressed air system is constructed, specifically including the following steps:

[0012] S11, acquire structural parameters of an air compression station house and a pipe network layout, and construct a three-dimensional model of the compressed air system;

[0013] S12, acquire a mechanism model of the compressed air system, and construct a digital twin of the compressed air system based on the three-dimensional model and the mechanism model;

[0014] S13, acquire data of the compressed air system in real time, and drive corresponding components of the digital twin in real time.

[0015] Preferably, the mechanism model comprises an air compressor model and a pipe network model.

[0016] Preferably, in step S2, the establishing a gas consumption prediction model according to the real-time data and the historical data specifically comprises: establishing a gas consumption prediction model by using a long short-term memory network algorithm, training the real-time data and the historical data as training samples of the gas consumption prediction model, and obtaining the gas consumption prediction model.

[0017] Preferably, in step S3, the constraint conditions comprise maintaining the pipe network pressure in a set range, an interval time between starting and stopping of a single air compressor >600 seconds, and a loading rate of the air compressor <40%.

[0018] Preferably, in step S4, the target function is solved by using a mixed integer programming or a deep reinforcement learning algorithm.

[0019] Preferably, the method further comprises: detecting and positioning a leak of the compressed air system based on a minimum flow method at night.

[0020] Preferably, the method further comprises: acquiring time sequence features of air compressor operation data, and performing maintenance on the air compressor based on the time sequence features by using a predictive maintenance model.

[0021] The application also provides a compressed air system energy-saving control system, which adopts the compressed air system energy-saving control method as described above, and comprises:

[0022] A digital twin construction module is configured to acquire various parameters of the compressed air system, and construct a digital twin of the compressed air system.

[0023] A model establishment module is configured to acquire real-time data and historical data of the compressed air system based on the digital twin, establish a gas consumption prediction model according to the real-time data and the historical data, and predict gas consumption demand in a future period based on the gas consumption prediction model.

[0024] A target function establishment module is configured to establish a target function with the minimum total energy consumption as the target on the premise of meeting the gas consumption demand in the future period, and set constraint conditions of the target function.

[0025] a target function solving module configured to solve the target function to obtain an optimal control strategy of the compressed air system that satisfies the constraint condition;

[0026] a control module configured to control the compressed air system based on the optimal control strategy.

[0027] Preferably, the digital twin construction module further comprises:

[0028] a three-dimensional model construction unit configured to obtain structural parameters of an air compression station and a pipe network layout, and construct a three-dimensional model of the compressed air system;

[0029] a digital twin construction unit configured to obtain a mechanism model of the compressed air system, and construct a digital twin of the compressed air system based on the three-dimensional model and the mechanism model;

[0030] a digital twin driving unit configured to collect data of the compressed air system in real time, and drive corresponding components of the digital twin in real time.

[0031] Compared with the prior art, the method has the following beneficial effects:

[0032] The method of the present application realizes visual monitoring of the compressed air system by constructing a digital twin of the compressed air system, and provides a high-precision virtual model and real-time data for subsequent use. A gas consumption prediction model is established based on real-time data and historical data provided by the digital twin, and the gas consumption prediction model is used to predict gas consumption in a future period. Under the premise of meeting the gas consumption demand in the future period, a target function is established with the objective of minimizing total energy consumption, and constraint conditions of the target function are set. The target function is solved to obtain an optimal control strategy of the compressed air system that satisfies the constraint condition. The compressed air system is controlled based on the optimal control strategy, so that the compressed air system effectively reduces energy consumption while meeting the gas consumption demand. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 A flowchart of the method for energy-saving control of a compressed air system. DETAILED DESCRIPTION

[0034] The technical solutions and technical effects of the embodiments of the present application are further described in detail below in combination with the accompanying drawings of the present application.

[0035] Please refer to Figure 1 A method for energy-saving control of a compressed air system, comprising.

[0036] S1, obtaining parameters of the compressed air system, and constructing a digital twin of the compressed air system;

[0037] S2, obtain real-time data and historical data of the compressed air system based on the digital twin, establish a gas consumption prediction model according to the real-time data and the historical data, and predict future period gas consumption based on the gas consumption prediction model;

[0038] S3, under the premise of meeting the future period gas consumption demand, a target function is established with the target of minimizing the total energy consumption, and the constraint condition of the target function is set;

[0039] S4, the target function is solved to obtain the optimal control strategy of the compressed air system meeting the constraint condition;

[0040] S5, the compressed air system is controlled based on the optimal control strategy.

[0041] The present application realizes the visual monitoring of the compressed air system by constructing the digital twin of the compressed air system, provides high-precision virtual models and real-time data for subsequent use, establishes a gas consumption prediction model based on the real-time data and historical data provided by the digital twin, and predicts future period gas consumption using the gas consumption prediction model. Under the premise of meeting the future period gas consumption demand, a target function is established with the target of minimizing the total energy consumption, and the constraint condition of the target function is set; the target function is solved to obtain the optimal control strategy of the compressed air system meeting the constraint condition; the compressed air system is controlled based on the optimal control strategy, so that the compressed air system meets the gas consumption demand while effectively reducing the energy consumption of the compressed air system.

[0042] Further, in step S1, the parameters of the compressed air system are obtained, and the digital twin of the compressed air system is constructed, specifically including the following steps:

[0043] S11, obtain the structural parameters of the air compression station house and the pipe network layout, and construct a three-dimensional model of the compressed air system; specifically, the compressed air system includes a plurality of devices, and in the present application, the air compressor unit of the gas supply end and the air pipe network of the transmission end of the compressed air system are mainly concerned. First, the three-dimensional data of the air compression station house and the pipe network layout in the physical world are obtained through sensors, cameras and other devices and drawings, then the collected data is cleaned, fused and optimized, and a high-precision three-dimensional model is generated through an algorithm, finally the three-dimensional model is rendered using a 3D engine, so that the three-dimensional model of the compressed air system is presented as a visual effect.

[0044] S12, acquire a mechanism model of the compressed air system, and construct a digital twin of the compressed air system based on the three-dimensional model and the mechanism model; the mechanism model is an abstract description of the operation mechanism or physical law of a physical device, and through the explicit mechanism model, the behavior or state of the physical object can be predicted through the virtual body. The calculation process and logic of the mechanism model are embedded into the corresponding components or objects of the three-dimensional model of the compressed air system, and the mechanism model is called for real-time calculation and simulation during the operation of the three-dimensional model, to obtain the digital twin of the compressed air system.

[0045] S13, real-time acquisition of data of the compressed air system, and real-time driving of corresponding components of the digital twin. The data of the compressed air system is acquired in real time through various sensors of the compressed air system, the real-time acquired data of the compressed air system is compared with the calculation result of the mechanism model, and the corresponding components of the digital twin are driven according to the comparison result, so that the physical entity and the virtual entity of the compressed air system are real-time symbiotic, and the compressed air which is originally "invisible and intangible" becomes all-process visual, measurable, controllable and optimal.

[0046] Further, the mechanism model includes an air compressor model and a pipe network model. The air compressor model adopts a nonlinear relationship model, the input of the model is the suction temperature, pressure and output pressure of the air compressor, and the output of the model is the specific power, exhaust volume and current of the air compressor. The pipe network model is a mathematical model of pressure-flow-pipe diameter based on fluid mechanics, used for simulating the pressure transmission loss of the pipe network.

[0047] Further, in step S2, a gas consumption prediction model is established according to real-time data and historical data, specifically including: a long short-term memory network algorithm is used to establish a gas consumption prediction model, real-time data and historical data are used as training samples of the gas consumption prediction model for training, and the gas consumption prediction model is obtained. The present application adopts a bidirectional long short-term memory network, inputs real-time data and historical data of the compressed air system in the input layer, and predicts the gas consumption demand in the future period in the output layer. The real-time data includes time data, production plan scheduling and environmental data; the time data includes the specific time currently acquired, the current running time, and whether the current is a working day or a weekend; the production plan scheduling refers to the planned arrangement of each production link on the production line, including the production time point, product category and production time length; the environmental data is the current outdoor temperature; the historical data includes the historical gas consumption records of the air compressors in each workshop.

[0048] Further, in step S3, the constraint conditions include that the pipe network pressure is maintained in a set range, the single air compressor start-stop interval time is >600 seconds, and the air compressor loading rate is <40%. The pipe network pressure is maintained at 0.65±0.02 MPa.

[0049] Further, in step S4, the objective function is solved by using mixed integer programming or deep reinforcement learning algorithm. Specifically, when the objective function is solved by using mixed integer programming, a mixed integer programming model needs to be constructed, including defining variables, constructing the objective function, setting constraint conditions, wherein the air compressor start-stop (0 / 1 variable), the loading state (continuous variable) are taken as optimization variables, the foregoing objective function is taken as the objective function of the model, and the foregoing constraint conditions are taken as the constraint conditions of the model. The optimal combination is solved once in each control period (such as 5 minutes), which is the optimal control strategy to be solved, including the start-stop of the air compressor and the loading state. When the objective function is solved by using deep reinforcement learning algorithm, the AI learns by itself through continuous interaction with the environment (executing control and observing energy saving effect), and finally obtains the optimal control strategy.

[0050] In the compressed air system, in addition to the lack of accurate prediction of air demand and dynamic response mechanism, which leads to a large amount of energy waste, the leakage of the compressed air system is a common energy waste source in the industrial field that is often ignored. According to research, the leakage of compressed air in the valve, joint, three-way joint, solenoid valve and cylinder head of the air-consuming equipment in the factory usually accounts for 10%-30% of the air supply. The loss caused by leakage not only directly wastes energy, but also causes the system pressure to drop. In order to ensure production demand, the air compressor has to increase the output pressure or increase the running time, further increasing the energy consumption. Therefore, in the present application, the compressed air system also needs to be detected and positioned for leakage, so as to reduce the energy waste caused by leakage.

[0051] In some embodiments, the method further comprises: performing leakage detection and positioning on the compressed air system based on the minimum night flow method. Since the abnormal increase in pipeline flow during the off-peak period is mainly caused by leakage, the leakage area and degree can be identified by analyzing the flow data during this period. Specifically, the compressed air system automatically identifies the off-peak period, which is generally 1:00-5:00 in the morning. During this period, the total flow of the pipeline is obtained in real time through the digital twin. If the total flow of the pipeline is continuously higher than the threshold value, such as 5% of the total air supply capacity, it is determined that there is a significant leakage in the compressed air system. Then, the leakage area needs to be located. Specifically, the flow data collected by each workshop branch flow sensor is obtained, and the collected flow data is compared with the preset flow data corresponding to each workshop branch. If the flow data collected by a branch is abnormally high compared with the preset flow data corresponding to the branch, it is determined that the branch is a leakage area. Since each branch flow sensor has a corresponding number, the specific area and position of the branch flow sensor can be found in the digital twin according to the number, so that the specific area of the leakage can be accurately located in the digital twin, and the specific position of the leakage area in the real world can be determined, so that the staff can check and repair in time, and reduce energy waste caused by leakage.

[0052] Further, predictive maintenance can also be performed on the compressed air system to timely detect faults in the compressed air system and provide stable pressure supply for the single crystal furnace. In some embodiments, the method further comprises: obtaining time sequence features of air compressor operation data, and performing maintenance on the air compressor based on the time sequence features using a predictive maintenance model. Historical normal operation data and historical fault data are obtained, wherein the historical normal operation data includes time sequence features of normal operation data and corresponding normal states, and the historical fault data includes time sequence features of fault data and corresponding fault conditions. For example, when the time sequence feature is an increase in current harmonic distortion rate, the corresponding fault condition is that the motor bearing may be worn out; when the time sequence feature is an abnormal decrease in the temperature difference between the exhaust temperature and the cooling water temperature, the corresponding fault condition is that the cooler may be fouled; and when the time sequence feature is a same-period extension of the loading time, the corresponding fault condition is a decrease in the unit efficiency, which may indicate an internal leakage. The classification model is trained using the historical normal operation data and the historical fault data to obtain the predictive maintenance model, which is used to predict the current state of the compressed air system. Thus, through predictive maintenance, passive repair is changed to active maintenance, which greatly reduces the risk of production stoppage caused by sudden faults in the air compressor, and provides stable pressure supply for the single crystal furnace to improve product yield.

[0053] The application also provides a compressed air system energy-saving control system, which adopts the compressed air system energy-saving control method described above, and comprises:

[0054] A digital twin construction module is configured to acquire parameters of the compressed air system and construct a digital twin of the compressed air system.

[0055] A model establishment module is configured to acquire real-time data and historical data of the compressed air system based on the digital twin, establish a gas consumption prediction model according to the real-time data and the historical data, and predict gas consumption in a future period based on the gas consumption prediction model.

[0056] A target function establishment module is configured to establish a target function with the minimum total energy consumption as the target under the premise of meeting the gas consumption in the future period, and set a constraint condition of the target function.

[0057] A target function solving module is configured to solve the target function to obtain an optimal control strategy of the compressed air system that meets the constraint condition.

[0058] A control module is configured to control the compressed air system based on the optimal control strategy.

[0059] Further, the digital twin construction module further comprises:

[0060] A three-dimensional model construction unit is configured to acquire structural parameters of an air compression station house and a pipe network layout, and construct a three-dimensional model of the compressed air system.

[0061] A digital twin construction unit is configured to acquire a mechanism model of the compressed air system, and construct a digital twin of the compressed air system based on the three-dimensional model and the mechanism model.

[0062] A digital twin driving unit is configured to acquire data of the compressed air system in real time, and drive corresponding components of the digital twin in real time.

[0063] The above only discloses preferred embodiments of the present application, and of course cannot limit the scope of the present application. Those skilled in the art can understand that all or part of the above-mentioned embodiments can be implemented, and equivalent changes made according to the claims of the present application still fall within the scope of the present application.

Claims

1. A method of energy saving control of a compressed air system, characterized in that, The method comprises the following steps: S1, acquiring parameters of a compressed air system, and constructing a digital twin of the compressed air system; S2, acquiring real-time data and historical data of the compressed air system based on the digital twin, establishing a gas consumption prediction model according to the real-time data and the historical data, and predicting gas consumption in a future period based on the gas consumption prediction model; S3, under the premise of meeting the gas consumption demand in the future period, establishing a target function with the objective of minimizing total energy consumption, and setting constraint conditions of the target function; S4, solving the target function to obtain an optimal control strategy of the compressed air system that meets the constraint conditions; S5, controlling the compressed air system based on the optimal control strategy.

2. The compressed air system energy saving control method of claim 1, wherein, In step S1, the parameters of the compressed air system are acquired, and the digital twin of the compressed air system is constructed, specifically comprising the following steps: S11, acquiring structural parameters of an air compression station house and pipe network layout, and constructing a three-dimensional model of the compressed air system; S12, acquiring a mechanism model of the compressed air system, and constructing a digital twin of the compressed air system based on the three-dimensional model and the mechanism model; S13, real-time acquisition of data of the compressed air system, and real-time driving of corresponding components of the digital twin.

3. The compressed air system energy saving control method of claim 2, wherein, The mechanism model comprises an air compressor model and a pipe network model.

4. The compressed air system energy saving control method of claim 1, wherein, In step S2, the gas consumption prediction model is established by using a long short-term memory network algorithm, the real-time data and the historical data are taken as training samples of the gas consumption prediction model for training, and the gas consumption prediction model is obtained.

5. The compressed air system energy saving control method of claim 1, wherein, In step S3, the constraint conditions comprise maintaining the pipe network pressure in a set range, an interval time between starting and stopping of a single air compressor >600 seconds, and a loading rate of the air compressor <40%.

6. The compressed air system energy saving control method of claim 5, wherein, In step S4, the target function is solved by using a mixed integer programming or deep reinforcement learning algorithm.

7. The compressed air system energy saving control method of claim 1, wherein, The method further comprises: performing leakage detection and positioning on the compressed air system based on a minimum flow method at night.

8. The compressed air system energy savings control method of claim 1, wherein, The method further comprises: acquiring time sequence characteristics of air compressor operation data, and performing maintenance on the air compressor based on the time sequence characteristics by using a predictive maintenance model.

9. A compressed air system energy saving control system employing the compressed air system energy saving control method according to any one of claims 1 to 8, characterized by The method comprises: a digital twin construction module configured to acquire parameters of a compressed air system, and construct a digital twin of the compressed air system; a model establishment module configured to acquire real-time data and historical data of the compressed air system based on the digital twin, establish a gas consumption prediction model according to the real-time data and the historical data, and predict gas consumption in a future period based on the gas consumption prediction model; a target function establishment module configured to establish a target function with the objective of minimizing total energy consumption under the premise of meeting the gas consumption demand in the future period, and set constraint conditions of the target function; a target function solving module configured to solve the target function to obtain an optimal control strategy of the compressed air system that meets the constraint conditions; a control module configured to control the compressed air system based on the optimal control strategy.

10. The compressed air system energy savings control system of claim 9, wherein, The digital twin construction module further comprises: A three-dimensional model construction unit is configured to acquire structural parameters of an air compression station house and a pipe network layout, and construct a three-dimensional model of the compressed air system; A digital twin construction unit is configured to acquire a mechanism model of the compressed air system, and construct a digital twin of the compressed air system based on the three-dimensional model and the mechanism model; A digital twin driving unit is configured to acquire data of the compressed air system in real time, and drive corresponding components of the digital twin in real time.

Images