Machine learning based refrigeration equipment energy saving control system and energy saving refrigeration equipment

Abstract

The application relates to the technical fields of machine learning and energy-saving refrigeration, and provides a refrigeration equipment energy-saving control system based on machine learning and an energy-saving refrigeration equipment, which is characterized in that a filtering square pipe, an air inlet square pipe, a sensor module and a machine learning module are arranged, two independent clean channels and an air filter screen with an adjustable inclination angle are arranged in the filtering square pipe, a selection component is arranged in the air inlet square pipe, and the selection component can be switched to any clean channel to form a filtering channel. The sensor module comprises air volume, pressure difference and dust concentration sensors, can monitor data in real time and transmit the data to the machine learning module. The machine learning module can predict the best switching time, the air filter screen inclination angle adjustment requirement, the filter screen cleaning requirement and the refrigeration load adjustment requirement according to the monitoring data, output corresponding control signals, complete filtering channel switching, air filter screen angle adjustment, filter screen cleaning and refrigeration load adjustment, and realize energy saving, cost reduction and equipment life improvement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical fields of machine learning, energy-saving refrigeration, and specifically relates to a refrigeration equipment energy-saving control system based on machine learning and an energy-saving refrigeration equipment. BACKGROUND

[0002] The existing refrigeration equipment generally relies on traditional filtering systems and constant load control strategies. In actual use, due to the fixed installation of the air filter screen in the filtering device all year round, it cannot be dynamically adjusted according to multi-dimensional factors such as actual air volume, dust concentration, pressure difference, etc., resulting in low filtering efficiency, serious energy waste, high maintenance cost, shortened filter screen life, and other problems. For example, when the air filter screen gradually accumulates dust, the air resistance increases, and the air volume correspondingly decreases. In order to maintain the refrigeration effect, it is usually necessary to increase the fan speed or increase the refrigeration load, so that the overall energy consumption of the system further rises. The traditional filtering structure is difficult to realize online switching or air filter screen inclination adjustment during use, and there is no effective pressure difference and dust monitoring to determine the cleaning time, resulting in high equipment energy consumption and maintenance cost. In the absence of effective monitoring and prediction, the filter screen often cannot be cleaned for a long time or can only be cleaned regularly by manual experience. Excessive dust accumulation may cause the filter screen to deform or perforate, thereby requiring replacement of the filter screen, resulting in continuous increase in maintenance cost. If cleaning is not timely or not thorough, it may also damage the normal operation of the system and shorten the overall service life of the equipment.

[0003] Therefore, how to reduce the energy consumption of the refrigeration equipment, improve the filtering efficiency, reduce the maintenance cost, and prolong the service life of the equipment has become a technical problem to be solved. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a refrigeration equipment energy-saving control system based on machine learning and an energy-saving refrigeration equipment to adaptively control the filtering channel switching, air filter screen inclination adjustment and cleaning, and refrigeration load adjustment, so as to realize energy saving and cost reduction and prolong the service life of the equipment.

[0005] In a first aspect, the present application provides a refrigeration equipment energy-saving control system based on machine learning, comprising:

[0006] A filter square pipe is provided with two structure-symmetrical and independent clean channels, and an air filter screen with an adjustable inclination angle is installed in each clean channel.

[0007] An air inlet square pipe is connected to the air inlet end of the filter square pipe, and a selection component for switching the communication between the air inlet square pipe and any clean channel to form a filtering channel is arranged in the air inlet square pipe.

[0008] A sensor module, which comprises an air volume sensor, a differential pressure sensor and a dust concentration sensor respectively installed in the air inlet square tube and the clean channel, the air volume sensor is used to monitor air volume data in real time, the differential pressure sensor is used to monitor differential pressure data in real time, and the dust concentration sensor is used to monitor dust data in real time;

[0009] A machine learning module, which is connected in communication with the sensor module, receives monitoring data monitored by the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time, predicts switching time for switching to form the filtering channel, air filter screen inclination angle adjustment demand, air filter screen cleaning demand and refrigeration load adjustment demand according to the monitoring data, outputs corresponding control signals, controls the selection assembly to switch the filtering channel, controls air filter screen inclination angle adjustment, air filter screen cleaning and refrigeration load adjustment.

[0010] In the second aspect, the application provides an energy-saving refrigeration equipment, which uses a machine learning-based refrigeration equipment energy-saving control system to perform energy-saving refrigeration.

[0011] Compared with the prior art, the application has the following beneficial effects:

[0012] The application provides a machine learning-based refrigeration equipment energy-saving control system and an energy-saving refrigeration equipment, which are characterized in that a filtering square tube, an air inlet square tube, a sensor module and a machine learning module are arranged, two clean channels which are structurally symmetrical and independent of each other are arranged in the filtering square tube, an air filter screen with an adjustable inclination angle is arranged in each clean channel, the air inlet square tube is connected to an air inlet end of the filtering square tube, a selection assembly for switching the air inlet square tube and any clean channel to form a filtering channel is arranged in the air inlet square tube, the sensor module comprises an air volume sensor, a differential pressure sensor and a dust concentration sensor respectively installed in the air inlet square tube and the clean channel, the air volume sensor is used to monitor air volume data in real time, the differential pressure sensor is used to monitor differential pressure data in real time, and the dust concentration sensor is used to monitor dust data in real time, the machine learning module is connected in communication with the sensor module, receives monitoring data monitored by the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time, predicts switching time for switching to form the filtering channel, air filter screen inclination angle adjustment demand, air filter screen cleaning demand and refrigeration load adjustment demand according to the monitoring data, outputs corresponding control signals, controls the selection assembly to switch the filtering channel, controls air filter screen inclination angle adjustment, air filter screen cleaning and refrigeration load adjustment, thereby realizing adaptive control of filtering channel switching, air filter screen inclination angle adjustment and cleaning, refrigeration load adjustment, achieving energy saving and cost reduction, and prolonging the service life of the equipment. BRIEF DESCRIPTION OF DRAWINGS

[0013] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0014] Figure 1 is a schematic diagram of a system architecture of a machine learning-based energy-saving control system for a refrigeration device according to an embodiment of the application. DETAILED DESCRIPTION

[0015] In order to enable persons skilled in the art to better understand the application scheme, the technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by persons skilled in the art without creative work should fall within the protection scope of the application.

[0016] Referring to Figure 1 , the embodiment of the application provides a machine learning-based energy-saving control system for a refrigeration device, comprising:

[0017] The filter square pipe is internally provided with two clean channels which are symmetrical in structure and independent of each other, and an adjustable-tilt-angle air filter screen is installed in each clean channel.

[0018] The air inlet square pipe is connected to the air inlet end of the filter square pipe, and a selection component for switching the communication between the air inlet square pipe and any clean channel to form a filter channel is arranged in the air inlet square pipe.

[0019] The sensor module comprises an air volume sensor, a differential pressure sensor and a dust concentration sensor which are respectively installed in the air inlet square pipe and the clean channels, the air volume sensor is used for monitoring air volume data in real time, the differential pressure sensor is used for monitoring differential pressure data in real time, and the dust concentration sensor is used for monitoring dust data in real time.

[0020] The machine learning module is in communication connection with the sensor module, receives monitoring data monitored by the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time, predicts switching timing of the filter channel, air filter screen inclination adjustment demand, air filter screen cleaning demand and refrigeration load adjustment demand according to the monitoring data, and outputs corresponding control signals to control the selection assembly to switch the filter channel, control air filter screen inclination adjustment, air filter screen cleaning and refrigeration load adjustment.

[0021] It should be noted that, in order to improve the energy efficiency of the refrigeration equipment under different operating conditions, the filter channel needs to have dynamic adjustment and adaptive switching to meet various air volume and dust concentration conditions. In addition, the traditional refrigeration system is difficult to accurately predict the optimal adjustment timing of the filter screen according to real-time monitoring data, resulting in excessive energy consumption or reduced filtering effect. In the embodiment, the selection assembly is configured in the air inlet square pipe to freely switch between two independent channels, reducing the waste caused by single filter screen blockage or insufficient efficiency. The air volume sensor, the differential pressure sensor and the dust concentration sensor of the sensor module can obtain key data in real time to provide accurate input for the machine learning module. The machine learning module dynamically predicts and outputs control signals according to the sensor information to coordinate the switching of the filter channel, the adjustment of the filter screen angle, the cleaning and the matching of the refrigeration load, thereby realizing the comprehensive goals of adaptability, high efficiency and energy saving, and effectively solving the problems of traditional refrigeration equipment filtering systems, such as difficulty in balancing high filtering efficiency and low energy consumption, and difficulty in realizing intelligent scheduling.

[0022] In some preferred embodiments, the selection assembly includes a rotatable assembly column installed in the air inlet square pipe and a gas guide groove arranged on the assembly column; the assembly column changes the communication position of the gas guide groove with one of the two clean channels by rotating to form the filter channel and realize the switching of the air flow path between the air inlet square pipe and the two clean channels. It should be noted that, in the embodiment, a rotatable assembly column and a gas guide groove are arranged in the air inlet square pipe to change the communication position of the gas guide groove with the clean channel by rotating the assembly column to complete the switching of the filter channel. When the filter screen is heavily dusted or the filtering efficiency is reduced, if the original channel is still forcibly maintained, it will cause the energy consumption to rise and the refrigeration effect to deteriorate. By designing a rotatable assembly column in the air inlet square pipe, the system can quickly switch between the two clean channels to avoid performance degradation caused by long-term use of the same filtering path. At the same time, the structural design of the assembly column and the gas guide groove can be flexibly arranged in a limited space, making the channel switching more concise, the sealing better and the air flow resistance smaller. The embodiment not only improves the operating efficiency of the equipment, but also automatically selects the optimal channel to ensure the filtering effect and reduce the overall energy consumption by combining the prediction results of the machine learning, thereby effectively solving the high energy consumption problem caused by the lack of flexible switching mechanism in the traditional refrigeration system.

[0023] In some preferred embodiments, the machine learning module predicts, according to the differential pressure monitored by the differential pressure sensor, the dust concentration monitored by the dust concentration sensor, and the air volume monitored by the air volume sensor, when the filtering effect of the clean channel in use is reduced or the energy consumption is significantly increased, outputs a filter channel switching control signal to control the assembly column to rotate and switch the filter channel. It should be noted that in this embodiment, the machine learning module uses the real-time data of the differential pressure sensor, the dust concentration sensor, and the air volume sensor to predict whether the filtering effect of the channel in use is reduced. If the energy consumption is significantly increased or the filtering effect is reduced, a channel switching control signal is output. The traditional system mainly relies on fixed time or manual experience to determine whether the filter screen needs to be replaced or switched, and it is difficult to accurately identify when the filtering efficiency is significantly reduced or the energy consumption is significantly increased. With the aid of multi-dimensional data of differential pressure, dust concentration, and air volume, machine learning can comprehensively judge the filter screen clogging trend and filter screen load, thereby triggering channel switching at the best time. This not only reduces the additional energy consumption caused by excessive clogging of the filter screen, but also effectively avoids resource waste caused by frequent switching, allowing the system to operate at the best working condition, greatly improving energy efficiency and extending the service life of the filter screen.

[0024] In some preferred embodiments, the machine learning module predicts, according to the dust concentration data monitored by the dust concentration sensor and the air volume data monitored by the air volume sensor, the degree of improvement of air filtering efficiency caused by increasing the inclination angle of the air filter screen and the resulting change in air resistance, and when the comprehensive benefits of filtering efficiency improvement exceed the impact of air volume reduction caused by increased air resistance, outputs an angle inclination control signal to control the adjustment of the windward inclination angle of the air filter screen. It should be noted that in this embodiment, the machine learning module uses dust concentration and air volume data to predict whether increasing the inclination angle of the air filter screen helps improve filtering efficiency, and outputs a control signal to adjust the inclination angle when the benefits of filtering efficiency improvement outweigh the losses caused by air volume reduction due to increased air resistance. The inclination angle of the filter screen is an important factor affecting airflow distribution and dust deposition. If the inclination angle is too small, the filtering efficiency may be insufficient, and if the inclination angle is too large, the filtering effect may increase but the air resistance may also increase. In traditional designs, the inclination angle of the filter screen is often fixed, which not only cannot adapt to environmental changes, but also causes energy waste or insufficient filtration. In this embodiment, the machine learning uses real-time dust and air volume data to balance the benefits and losses, ensuring that the optimal inclination angle range is automatically found under different working conditions, reducing energy consumption and maintaining excellent filtering effect.

[0025] In some preferred embodiments, when the air volume sensor monitors that the real-time air volume is lower than a preset value and the dust concentration sensor monitors that the real-time dust concentration is lower than a set threshold value, the machine learning module outputs a control signal for reducing the windward tilt angle of the air filter screen, and controls the windward tilt angle of the air filter screen to be reduced. It should be noted that when the air volume monitoring value is lower than the preset value and the dust concentration is lower than the threshold value, the machine learning module outputs a control signal for reducing the windward tilt angle of the air filter screen. If the environmental dust concentration is not high and the air volume has dropped below a certain critical level, continuing to maintain a large tilt angle of the filter screen at this time will cause the air resistance to be too large, resulting in further decrease of the air volume, causing insufficient refrigeration efficiency or increased load of the compressor and the fan. Based on this, in the present embodiment, the system analyzes the sensor data in real time through machine learning, and if it is found that the dust concentration is low and does not need to be filtered strongly, the angle of the filter screen can be adjusted to be small to reduce the resistance, thereby saving energy and improving the overall efficiency of the equipment to the maximum extent on the premise of meeting the basic requirements of filtering performance.

[0026] In some preferred embodiments, the machine learning module analyzes the change trend of the differential pressure monitored by the differential pressure sensor, the change trend of the air dust concentration monitored by the dust concentration sensor, and the change trend of the air volume monitored by the air volume sensor in real time, predicts the clogging state of the air filter screen, and outputs an air filter screen cleaning control signal when the predicted clogging degree reaches a set value to control the air filter screen to be cleaned. It should be noted that in the present embodiment, the change trends of the three types of sensor data of differential pressure, dust concentration and air volume are comprehensively analyzed to predict the clogging state of the air filter screen, and the air filter screen cleaning control signal is output when the clogging degree reaches a set value. It can be understood that relying solely on artificial experience may result in the filter screen being cleaned only after being excessively clogged, which not only keeps the energy consumption at a high level for a long time, but also damages the equipment. If the cleaning is too frequent, the maintenance cost and downtime will be increased. In the present embodiment, by monitoring the change trends of the three types of sensor data in real time and using a machine learning algorithm, the system can more accurately judge the clogging degree, avoid the two extreme situations of excessive clogging and frequent cleaning, and achieve a balance between energy saving and stable operation.

[0027] In some preferred embodiments, a filter screen cleaner is arranged in the clean channel, and the filter screen cleaner cleans the air filter screen according to the air filter screen cleaning control signal. It should be noted that the filter screen cleaner is arranged in the clean channel, and the filter screen cleaner cleans the air filter screen according to the air filter screen cleaning control signal, which can realize online cleaning or semi-automatic cleaning in combination with the cleaning signal of the machine learning module, saving time and effort, and ensuring continuous and efficient operation of the equipment.

[0028] In some preferred embodiments, the machine learning module predicts the refrigeration load demand according to the monitoring data monitored by the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time, outputs a refrigeration load control signal, and controls the adjustment of the compressor frequency and the fan rotating speed. It should be noted that the main source of energy consumption of the refrigeration equipment is the compressor and the fan, and the traditional control mostly uses fixed frequency or simple PID algorithm, which is difficult to track the changes of external environment and filter system state in real time. If the filter resistance increases and the air volume decreases suddenly, the traditional strategy still maintains the fixed frequency, which may cause the refrigeration efficiency to decrease or the energy to be consumed excessively. In the present embodiment, the load is predicted and controlled in real time by the machine learning algorithm, so that the system operation can be more in line with the current demand, unnecessary energy loss can be reduced, and stable refrigeration effect can be maintained.

[0029] In some preferred embodiments, when the differential pressure monitored by the differential pressure sensor abnormally increases or the air volume monitored by the air volume sensor abnormally decreases by more than a preset value, the machine learning module predicts that the equipment may have a risk of failure, automatically outputs failure response control information, and controls the reduction of the compressor frequency or the fan rotating speed. It should be noted that in the refrigeration equipment, if the filter screen or the related pipeline is seriously blocked, extreme differential pressure increase and air volume decrease will occur, which will further cause equipment load abnormality, compressor or fan overload and other failure risks. If not handled in time, it will cause the equipment to shut down urgently or be damaged further. In the present embodiment, the risk is monitored and evaluated in real time by the machine learning, so that the control instruction can be issued in time before the failure occurs, the compressor frequency or the fan rotating speed is reduced, the high load risk is alleviated, and maintenance is reminded to prevent more serious loss.

[0030] In some preferred embodiments, the machine learning module uses a deep neural network model trained with the data of air volume, differential pressure and dust concentration in historical operation to realize filter channel switching prediction, filter screen inclination optimization, filter screen cleaning judgment and refrigeration load prediction. It should be noted that the traditional algorithm based on simple rules or linear model often cannot accurately capture the complex nonlinear relationship between multiple sensor data and equipment energy consumption and filtering efficiency; the deep neural network has stronger feature extraction capability and can learn more accurate decision rules under the training of massive historical data, so as to make more sensitive and reliable prediction and control on the changing state in operation, and enhance the self-learning and self-adaptive ability of the system.

[0031] In some preferred embodiments, the machine learning module is updated online during real-time operation according to the latest data collected by the air volume, pressure difference and dust concentration sensors, automatically adjusts the model parameters, and realizes self-adaptation to real-time working conditions and environmental changes. It should be noted that the environment of the refrigeration equipment (such as outdoor dust concentration, air temperature, humidity) may fluctuate greatly over time. If only relying on the model trained offline, the model accuracy may continue to decline as the equipment ages or the environment changes. The online training mechanism can enable the machine learning model to continuously learn new sensor data, thereby dynamically correcting the control strategy, maintaining high prediction accuracy and system efficiency, and avoiding control failure or mismatch problems caused by working condition deviation.

[0032] In some preferred embodiments, the input data of the machine learning module includes sliding window statistical features of air volume, pressure difference and dust concentration sensor data, and adopts random forest or XGBoost algorithm to perform prediction analysis on the real-time operation trend of the refrigeration equipment. It should be noted that the change of the refrigeration system state is not only related to the data at a single moment, but also closely related to the data trend in the past period of time. Using sliding window to extract data features can capture the time evolution law and reduce the influence of noise on instantaneous value. Random forest and XGBoost and other ensemble learning algorithms are good at processing multi-dimensional sensor data and non-linear relationships, and can mine more potential patterns in high-dimensional data, providing algorithm support for more accurate energy consumption prediction and control decisions of the system.

[0033] In some preferred embodiments, the machine learning module integrates a reinforcement learning algorithm, takes air volume, pressure difference and dust concentration data as real-time input, and optimizes the filter screen inclination angle, filter passage switching, filter screen cleaning timing and refrigeration load control strategy to optimize system energy consumption. It should be noted that reinforcement learning is suitable for handling control problems with trial-and-error mechanisms and long-term benefit optimization requirements. The energy consumption of the refrigeration equipment is not simply optimal at a certain moment, but also needs to consider multiple factors such as filter screen state change, cleaning frequency, equipment life, etc. Through reinforcement learning, the system can continuously accumulate experience and explore better behavior strategies in long-term operation, thereby achieving the goal of continuously optimizing energy consumption under different environmental and load conditions.

[0034] In some preferred embodiments, the machine learning module is deployed in a digital twin platform, in which a digital twin model is built, the machine learning module is deployed on the digital twin platform and is connected to the digital twin model; the digital twin model receives data of the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time to simulate the actual operation state of the refrigeration equipment, the machine learning module optimizes the control strategy based on the simulated operation data of the digital twin model, and the optimized control strategy is issued to the actual refrigeration equipment after being verified by the digital twin model in real time. It should be noted that the operating environment of the real refrigeration equipment is complex and changeable, and frequent parameter adjustment or trial and error directly on the entity equipment has high risk and high cost; while the digital twin model can receive sensor data in real time and perform high-precision simulation, provide a mirror environment for the machine learning module to verify and optimize the strategy, thereby reducing the interference and trial and error cost of the actual equipment and greatly shortening the optimization period.

[0035] In some preferred embodiments, the digital twin model receives data of the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time and performs dynamic simulation of the equipment operation state, the machine learning module predicts possible fault risks of the actual refrigeration equipment according to the simulation results of the digital twin model, and automatically generates maintenance warning information and maintenance suggestions, which are output to the control terminal of the actual refrigeration equipment. It should be noted that when the refrigeration equipment has subtle signs of failure, the sensor data and the digital twin model can capture the trend change earlier, and the machine learning model can make a prediction accordingly to avoid the expansion of the fault and cause serious loss. In this embodiment, by simulating various extreme situations in advance in the digital twin model, the machine learning algorithm can identify early features of the fault, and once similar early features of the fault are found, the warning information and the maintenance scheme can be output, thereby reducing the unplanned downtime, ensuring the stable operation of the equipment, and improving the foresight and accuracy of system maintenance.

Claims

1. A machine learning based energy saving control system for a refrigeration plant, characterized in that, The application relates to a filter device for air conditioning system, comprising: a filter square pipe, two symmetrical and independent clean channels are arranged in the filter square pipe, and an adjustable air filter screen with an adjustable inclination angle is arranged in each clean channel; an air inlet square pipe, the air inlet square pipe is connected with an air inlet end of the filter square pipe, and a selection assembly for switching the air inlet square pipe and any clean channel to form a filter channel is arranged in the air inlet square pipe; a sensor module, a wind volume sensor, a differential pressure sensor and a dust concentration sensor are arranged in the air inlet square pipe and the clean channel respectively, the wind volume sensor is used for monitoring wind volume data in real time, the differential pressure sensor is used for monitoring differential pressure data in real time, and the dust concentration sensor is used for monitoring dust data in real time; a machine learning module, which is connected with the sensor module, receives monitoring data monitored by the wind volume sensor, the differential pressure sensor and the dust concentration sensor in real time, predicts a switching time for forming the filter channel, an air filter screen inclination angle adjustment requirement, an air filter screen cleaning requirement and a refrigeration load adjustment requirement according to the monitoring data, and outputs corresponding control signals to control the selection assembly to switch the filter channel, control the air filter screen inclination angle adjustment, the air filter screen cleaning and the refrigeration load adjustment; the machine learning module predicts an increase degree of air filter screen inclination angle on the improvement of air filtering efficiency and a change of air resistance caused by the increase according to dust concentration data monitored by the dust concentration sensor in real time and wind volume data monitored by the wind volume sensor in real time, and outputs an angle inclination control signal to control the adjustment of the air filter screen windward inclination angle when the comprehensive benefits of the filtering efficiency improvement exceed the influence of the wind volume decrease caused by the increase of the air resistance.

2. The machine learning based energy saving control system for refrigeration equipment as claimed in claim 1 wherein, the selection assembly comprises an assembly column arranged in the air inlet square pipe and rotatable and a gas guide groove arranged on the assembly column; the assembly column changes the communication position of the gas guide groove and one of the two clean channels by rotating to form the filter channel and realize the switching of the air flow path between the air inlet square pipe and the two clean channels.

3. The machine learning based energy saving control system for refrigeration equipment according to claim 2, wherein, the machine learning module outputs a filter channel switching control signal to control the rotation of the assembly column to switch the filter channel when the filtering effect of the currently used clean channel is reduced or the energy consumption is obviously increased according to the differential pressure of the air filter screen before and after the clean channel monitored by the differential pressure sensor in real time, the air dust concentration monitored by the dust concentration sensor in real time and the wind volume data monitored by the wind volume sensor in real time.

4. The machine learning based energy saving control system for a refrigeration equipment as claimed in claim 1 wherein, the machine learning module outputs a control signal for reducing the air filter screen windward inclination angle to control the reduction of the air filter screen windward inclination angle when the real-time wind volume monitored by the wind volume sensor is lower than a preset value and the real-time dust concentration monitored by the dust concentration sensor is lower than a set threshold.

5. The machine learning based energy saving control system for a refrigeration equipment as claimed in claim 1 wherein, the machine learning module analyzes the differential pressure change trend of the air filter screen before and after the clean channel monitored by the differential pressure sensor, the air dust concentration change trend monitored by the dust concentration sensor and the wind volume change trend monitored by the wind volume sensor in real time, predicts the clogging state of the air filter screen, and outputs an air filter screen cleaning control signal to control the cleaning of the air filter screen when the predicted clogging degree reaches a set value.

6. The machine learning based energy saving control system for a refrigeration equipment as claimed in claim 5 wherein, The clean channel is provided with a filter screen cleaner, which cleans the empty filter screen according to the empty filter screen cleaning control signal.

7. The machine learning based energy saving control system for a refrigeration equipment as claimed in claim 1 wherein, The machine learning module predicts the refrigeration load demand according to the monitoring data monitored by the air volume sensor, the differential pressure sensor and the dust concentration sensor in real time, outputs a refrigeration load control signal, and controls the frequency of the compressor and the rotating speed of the fan.

8. The machine learning based energy saving control system for a refrigeration equipment as claimed in claim 7 wherein, When the differential pressure monitored by the differential pressure sensor abnormally rises or the air volume monitored by the air volume sensor abnormally decreases by more than a preset value, the machine learning module predicts that the equipment may have a risk of failure, automatically outputs failure response control information, and controls the compressor frequency or the rotating speed of the fan to be reduced.

9. An energy saving refrigeration apparatus, characterized by comprising: The energy-saving refrigeration equipment uses the machine learning-based refrigeration equipment energy-saving control system of any one of claims 1-8 for energy-saving refrigeration.

Images