Energy consumption control and regulation method and system for smart farm

Abstract

The application discloses an energy consumption control and regulation method and system for a smart farm, and relates to the technical field of intelligent regulation and control.The method comprises the following steps: acquiring historical temperature and humidity regulation values, historical light intensity regulation values and historical concentration regulation values; determining whether three types of data are all in the same regulation time point for environment regulation; counting three types of historical regulation schemes; training a simulation prediction model one; if the external environment characteristics of each historical period in the historical regulation record data present an unstable state, then counting a first variable regulation scheme and a second variable regulation scheme performed in a time period before a historical variable critical time point and a time period before and after the historical variable critical time point, training a simulation prediction model two, and testing the current prediction regulation scheme by the simulation prediction model one or the simulation prediction model two according to the current external environment characteristic value.The energy consumption control and regulation method and system for the smart farm can improve the energy consumption control efficiency.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology, and in particular to energy consumption control and regulation methods and systems for smart farms. Background Technology

[0002] Intelligent regulation of environmental control and energy consumption in smart farms is of great significance for improving breeding efficiency, reducing energy consumption, improving management efficiency, ensuring animal health, meeting environmental protection requirements, and optimizing production decisions. It is one of the key technologies for promoting the intelligent, efficient, and sustainable development of the breeding industry.

[0003] In existing technologies, the energy consumption caused by changes in the control parameters of equipment that respond to environmental factors in a timely manner often only considers the temperature, humidity, light intensity, and gas concentration in the environment and directly adjusts these parameters simultaneously. However, it does not analyze the different degrees of energy savings resulting from the different degrees of interrelationships between these three types of factors and the resulting diverse simultaneous control treatments, leading to low efficiency of the final energy consumption control scheme. Summary of the Invention

[0004] To overcome the shortcomings of the existing technology, this application provides a method and system for energy consumption control and regulation in smart farms.

[0005] Firstly, the energy consumption control and regulation method for a smart farm provided in this application includes:

[0006] Acquire historical control records of intelligent energy consumption control in historical periods. When three categories of data in the historical control records—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme one is calculated.

[0007] When two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control records belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme two is calculated. Alternatively, when three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental control at different control time points, then historical control scheme three is calculated.

[0008] Based on the historical control scheme 1, historical control scheme 2 and historical control scheme 3, a first training feature set, a second training feature set and a third training feature set are obtained. If the external environmental characteristics of each historical period in the historical control record data are stable, then a simulation prediction model 1 is trained based on the first training feature set, the second training feature set and the third training feature set.

[0009] If the external environmental characteristics of each historical period in the historical control record data are unstable, then from the historical change critical time point with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data, the first change control scheme implemented in the period before the historical change critical time point and the second change control scheme implemented in the period after the historical change critical time point are statistically determined. Based on the first change control scheme and the second change control scheme, simulation prediction model two is trained.

[0010] Obtain the current external environmental characteristic values ​​for the current period, determine whether the current external environmental characteristic values ​​are in a stable state, and select one of the simulation prediction models from simulation prediction model one and simulation prediction model two for testing to obtain the current prediction and control scheme.

[0011] Preferably, historical control records of intelligent energy consumption control in historical periods are obtained, and historical temperature and humidity control values, historical light intensity control values, and historical gas concentration control values ​​are extracted from the historical control records.

[0012] If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and environmental control is performed at the same control time point, then historical test case one will be output.

[0013] Based on the historical test case one, the first correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, historical light intensity control values ​​and historical concentration control values.

[0014] The historical test scenario 1 and the first correlation coefficient are combined to form historical control scheme 1.

[0015] Preferably, if two of the following categories of data belong to the same historical period and are subject to environmental regulation at the same regulation time point: historical test case two;

[0016] Based on the second historical test case, the second correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, the historical light intensity control values, and the historical concentration control values.

[0017] The second historical test scenario and the second correlation coefficient are combined to form the second historical control scheme.

[0018] If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and environmental control is carried out at different control time points, then output historical test case three;

[0019] Based on the aforementioned historical test case three, the third correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values.

[0020] The historical test scenario three and the third correlation coefficient are combined to form historical control scheme three.

[0021] Preferably, the actual energy consumption value after regulation was implemented in each historical period is extracted from the historical regulation record data, and the historical demand effect value before regulation was implemented in each historical period is extracted from the historical regulation record data.

[0022] The historical regulation scheme 1, the historical actual energy consumption value of regulation, and the historical demand effect value are combined to form the first training feature set;

[0023] The historical regulation scheme 2, the historical actual energy consumption value of regulation, and the historical demand effect value are combined to form the second training feature set;

[0024] The historical regulation scheme three, the historical regulation actual energy consumption value and the historical demand effect value are combined to form the third training feature set;

[0025] If the external environmental characteristics of each historical period in the historical regulation record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set.

[0026] Preferably, if the external environmental characteristics of each historical period in the historical control record data are in an unstable state, then the critical time point of the historical change with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data is selected.

[0027] If one of the historical control schemes 1, 2, and 3 is implemented within the time period before the critical point of the historical change, then the first change control scheme is output.

[0028] Furthermore, if a control scheme different from the first change control scheme is implemented among historical control scheme one, historical control scheme two, and historical control scheme three within the time period after the critical point of historical change, then the second change control scheme will be output.

[0029] A fourth correlation coefficient is obtained by extracting the degree of mutual influence between the first change control scheme implemented in the period before the critical point of historical change within the same historical period and the second change control scheme implemented in the period after the critical point of historical change.

[0030] Based on the first variable control scheme, the second variable control scheme, historical actual energy consumption values ​​and historical demand effect values, simulation prediction model two is trained.

[0031] Preferably, the current demand effect value and the current external environmental characteristic value that need to be controlled for energy consumption in the current period are obtained, and the demand environmental characteristic value that needs to be achieved in the current period is detected.

[0032] The difference between the current external environmental characteristic value and the required environmental characteristic value is calculated to obtain the preprocessed environmental difference value;

[0033] When it is determined that the current external environmental characteristic values ​​are in a stable state, the current demand effect value and the pre-processed environmental difference value are input into the simulation prediction model one for testing to obtain the current prediction and control scheme.

[0034] Alternatively, when it is determined that the current external environmental characteristic value is in an unstable state, the critical time point of the current change with the greatest degree of change in the external environmental characteristic is statistically identified.

[0035] The current demand effect value, preprocessing environment difference, and current critical time point are then input into simulation prediction model two for testing to obtain the current prediction and control scheme.

[0036] Secondly, an energy consumption control and regulation system for a smart farm includes:

[0037] Historical feature statistics unit one is used to acquire historical control records of intelligent energy consumption control in historical periods. When three categories of data in the historical control records—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and environmental control is carried out at the same control time point, then historical control scheme one is statistically calculated.

[0038] Historical feature statistics unit two is used to calculate historical control scheme two when two of the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control record data belong to the same historical period and are subject to environmental control at the same control time point; or, when three of the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental control at different control time points, historical control scheme three is calculated.

[0039] Model building unit one is used to obtain the first training feature set, the second training feature set, and the third training feature set based on historical control scheme one, historical control scheme two, and historical control scheme three. If the external environmental characteristics of each historical period in the historical control record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set.

[0040] Model building unit two is used to calculate the first change control scheme in the period before the historical change critical time point and the second change control scheme in the period after the historical change critical time point if the external environmental characteristics of each historical period in the historical control record data are unstable. Based on the first change control scheme and the second change control scheme, simulation prediction model two is trained.

[0041] The current control scheme prediction unit is used to obtain the current external environmental characteristic values ​​for the current period, determine whether the current external environmental characteristic values ​​are in a stable state, and select one of the simulation prediction models from simulation prediction model one and simulation prediction model two to test and obtain the current prediction control scheme.

[0042] Compared with the prior art, the present invention has the following characteristics and beneficial effects:

[0043] By conducting a differentiated analysis of various control schemes implemented under diverse conditions in historical control records, this study examines three types of control schemes. The first involves environmental control treatment based on historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​from historical control records. This analysis differentiates whether these three types of data belong to the same historical period and are implemented at the same control point. Since the degree of interrelationship between environmental impact factors resulting from simultaneous or asynchronous environmental control of these three types of data differs and directly reflects in energy consumption values, this differentiated analysis improves the rigor of handling the degree of correlation between multiple types of impact factors and enhances the efficiency of energy consumption control scheme implementation. Furthermore, the analysis assesses whether the external environmental characteristics are in a stable state. It further considers that if only one type of environmental control scheme is implemented when the external environmental characteristics are unstable, the energy-saving efficiency will not be high. Therefore, implementing two types of environmental control schemes can be considered to greatly improve the efficiency of energy conservation and utilization. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the steps of the energy consumption control and regulation method for a smart farm, which is the main feature of this embodiment.

[0045] Figure 2 This is a structural block diagram of an energy consumption control and regulation system for a smart farm, which is the main feature of this embodiment. Detailed Implementation

[0046] The present invention will be further described in detail below with reference to the following embodiments.

[0047] Reference Figure 1 A method for energy consumption control and regulation in smart farms, comprising the following steps:

[0048] S1. Obtain historical control records for intelligent energy consumption control during historical periods. When three categories of data in the historical control records—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme one is calculated.

[0049] S2. When two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control record data belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme two is calculated. Alternatively, when three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental control at different control time points, then historical control scheme three is calculated.

[0050] S3. If the external environmental characteristics of each historical period in the historical control record data are stable, then the first training feature set, the second training feature set, and the third training feature set are obtained according to the first historical control scheme, the second historical control scheme, and the third historical control scheme, respectively. Based on the first training feature set, the second training feature set, and the third training feature set, the simulation prediction model one is trained.

[0051] S4. If the external environmental characteristics of each historical period in the historical control record data are unstable, then from the historical change critical time point with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data, the first change control scheme implemented in the period before the historical change critical time point, and the second change control scheme implemented in the period after the historical change critical time point are statistically determined. Based on the first change control scheme and the second change control scheme, simulation prediction model two is trained.

[0052] S5. Obtain the current external environmental characteristic values ​​for the current period, determine whether the current external environmental characteristic values ​​are in a stable state, and select one of the simulation prediction models from simulation prediction model one and simulation prediction model two to test and obtain the current prediction and control scheme.

[0053] Specifically, a differentiated analysis was conducted on various control schemes implemented under diverse conditions in historical control records. These schemes included three types: first, environmental control treatment based on historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​from historical control records; second, a diversity-based analysis was performed on whether these three types of data belonged to the same historical period and were implemented at the same control point. Since the degree of interrelationship between environmental impact factors resulting from simultaneous or asynchronous environmental control of these three types of data differs and directly reflects in energy consumption values, this diversity-based analysis was conducted to improve the rigor of handling the degree of correlation between multiple types of impact factors and enhance the efficiency of energy consumption control scheme implementation. Third, the stability of external environmental characteristics was assessed. It was further considered that if only one type of environmental control scheme was implemented when external environmental characteristics were unstable, the energy-saving efficiency would not be high. Therefore, implementing two types of environmental control schemes could be considered to greatly improve the efficiency of energy conservation.

[0054] The specific step S1 includes the following sub-steps:

[0055] Historical control records of intelligent energy consumption regulation are obtained. From these records, historical temperature and humidity control values, light intensity control values, and gas concentration control values ​​are extracted.

[0056] If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and are subject to environmental control at the same control time point, then output historical test case one.

[0057] Based on the historical test case 1, the first correlation coefficient is obtained by extracting the degree of mutual influence coefficients among the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values.

[0058] The historical scenario to be tested and the first correlation coefficient are combined to form the historical control scheme one.

[0059] Specifically, historical control records (referring to environmental temperature and humidity data, light intensity data, and gas concentration data detected in the farm and external environment during the historical control period, as well as parameter values ​​of the control scheme and energy consumption values, etc.), and historical test cases one (if the environmental control of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​are respectively labeled A, B, and C for subsequent examples, the same historical period refers to, for example, the same day, and the same control time point refers to A and B). The implementation of schemes A, B, and C involves adjustments at a single point in time, such as simultaneous adjustments of various parameter values ​​at 10:00 AM. The first correlation coefficient (for example, if the theoretical parameter adjustments for A, B, and C are A1, B1, and C1 respectively, and the actual parameter adjustments for A, B, and C are A2, B2, and C2 respectively, the differences are a0 for A1-A2, b0 for B1-B2, and c0 for C1-C2. a0 / b0 is the correlation coefficient between A and B, and so on).

[0060] The specific step S2 includes the following sub-steps:

[0061] If two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and are subject to environmental control at the same control time point, then output historical test case two.

[0062] Based on the second historical test case, the second correlation coefficient is obtained by extracting the degree of mutual influence coefficients among the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values.

[0063] Historical scenario two and the second correlation coefficient are combined to form historical control scheme two.

[0064] If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and environmental control is carried out at different control time points, then the third historical test case will be output.

[0065] Based on the third historical test case, the correlation coefficient was obtained by extracting the degree of influence coefficients among the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values.

[0066] The historical scenario three and the third correlation coefficient are combined to form the historical control scheme three.

[0067] Specifically, there are historical scenarios to be tested, such as scenario two (which is explained the same as scenario one, but it should be noted that the two categories of data belong to the same historical period and are subject to environmental regulation at the same regulation time point: for example, A, B, and C are all regulated on the same historical day, but only the regulation of A and B is carried out at 10:00, while C is carried out before or after 10:00), scenario three (which is explained the same as scenario one, but it should be noted that the three categories of data belong to different historical periods and are subject to environmental regulation at different regulation time points: for example, A, B, and C are all regulated on the same historical day, but the regulation of A is carried out at 10:00, while B is carried out at 10:30, while C is carried out after 10:30, between 10:00 and 10:30, or before 10:00), the second correlation coefficient, and the third correlation coefficient (which are explained the same as the first correlation coefficient, and will not be explained further here).

[0068] The specific step S3 includes the following sub-steps:

[0069] The actual energy consumption values ​​after regulation were implemented in each historical period were extracted from the historical regulation record data, and the historical demand effect values ​​before regulation were implemented in each historical period were also extracted from the historical regulation record data.

[0070] The first training feature set is formed by combining the historical regulation scheme 1, the historical regulation actual energy consumption value, and the historical demand effect value.

[0071] The second training feature set is formed by combining the historical regulation scheme 2, the historical regulation actual energy consumption value and the historical demand effect value.

[0072] The third training feature set is formed by combining the historical regulation scheme three, the historical regulation actual energy consumption value and the historical demand effect value.

[0073] If the external environmental characteristics of each historical period in the historical regulation record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set.

[0074] Specifically, historical demand-effect values ​​(referring to the ideal energy consumption value (smaller than the actual energy consumption value in historical regulation)) and simulation prediction models (e.g., training separately based on the first, second, and third training feature sets; for example, the first training feature set: if the parameter values ​​for environmental regulation in historical regulation scheme one include three categories of data, such as s1, s2, and s3, then Y1 = (s1 + s2 + s3) * X1 + R1, where X1 refers to the first correlation coefficient and R1 refers to the corresponding historical regulation scheme). The first training feature set is the difference between the historical actual energy consumption value and the historical demand effect value. Y1 refers to the matching value corresponding to the historical control scheme one. Similarly, the second training feature set is: Y2 = (s1 + s2 + s3) * X2 + R2. It should be noted that s1, s2, and s3 here refer to s1, s2, and s3 in the historical control scheme two. The third training feature set is: Y3 = (s1 + s2 + s3) * X3 + R3. It should be noted that s1, s2, and s3 here refer to s1, s2, and s3 in the historical control scheme three.

[0075] The specific step S4 includes the following sub-steps:

[0076] If the external environmental characteristics of each historical period in the historical control record data are unstable, then the critical time point of the greatest change in the external environmental characteristics of each historical period in the historical control record data is selected.

[0077] If one of the historical control schemes (Scheme 1, Scheme 2, and Scheme 3) is implemented within the time period before the critical point of historical change, then the first change control scheme will be output.

[0078] Furthermore, if, within a time period following the critical point of historical change, a control scheme different from the first change control scheme is implemented among historical control scheme one, historical control scheme two, and historical control scheme three, then the second change control scheme will be output.

[0079] The fourth correlation coefficient is obtained by extracting the degree of mutual influence between the first change control scheme implemented in the period before the critical point of historical change within the same historical period and the second change control scheme implemented in the period after the critical point of historical change.

[0080] Based on the first and second variable control schemes, historical actual energy consumption values, and historical demand effect values, simulation prediction model two was trained.

[0081] Specifically, if the external environmental characteristics of each historical period in the historical control records are in an unstable state (meaning that as the values ​​of the external environmental characteristics change, the values ​​of the environmental control parameters will also rise or fall, and the control scheme implemented at this time is very likely to be significantly different from the one implemented when the external environmental characteristics are stable, thereby reducing energy consumption), the critical time point of historical change (e.g., if the change in the values ​​of the external environmental characteristics is stable before 5 PM, but the change in the values ​​of the external environmental characteristics is large after 5 PM, then 5 PM is the critical time point of historical change), the first change control scheme (e.g., if the environmental control scheme implemented before 5 PM is historical control scheme one, then it is the first change control scheme), the second change control scheme (e.g., if the environmental control scheme implemented after 5 PM is one of historical control scheme two or three, then it is the second change control scheme), and the fourth correlation coefficient (e.g., if the first change control scheme (if it is historical control scheme one)... Case 1): If the theoretical parameter adjustments for A, B, and C regulation are A1, B1, and C1 respectively, and the actual parameter adjustments for A, B, and C regulation are A2, B2, and C2 respectively, and the differences are a0 for A1-A2, b0 for B1-B2, and c0 for C1-C2, then a0 / b0 is the correlation coefficient between A and B. For the second change regulation scheme (if it is the second historical regulation scheme): if the correlation coefficient between A and B is a1 / b1, then (a0 / b0+a1 / b1) / 2 is the fourth correlation coefficient. Here, we take the correlation coefficient between A and B in the first and second change regulation schemes as an example for analysis. The rest are analogous and will not be explained in detail here. Simulation prediction model 2 (the explanation is the same as that of simulation prediction model 1. It should be noted that the first and second change regulation schemes include the historical change critical time point corresponding to the matching feature data).

[0082] The specific step S5 includes the following sub-steps:

[0083] Obtain the current demand effect value and the current external environmental characteristic value that require energy consumption regulation in the current period, and detect the demand environmental characteristic value that needs to be achieved in the current period.

[0084] The difference between the current external environmental characteristic values ​​and the required environmental characteristic values ​​is used to obtain the preprocessed environmental difference value.

[0085] When it is determined that the current external environmental characteristic values ​​are in a stable state, the current demand effect value and the pre-processed environmental difference value are input into the simulation prediction model 1 for testing to obtain the current prediction and control scheme.

[0086] Alternatively, when it is determined that the current external environmental characteristic value is in an unstable state, the critical time point of the current change with the greatest degree of change in the external environmental characteristic is statistically determined.

[0087] The current demand effect value, preprocessing environment difference, and current critical time point are then input into simulation prediction model two for testing to obtain the current prediction and control scheme.

[0088] Specifically, for example, pre-processing environmental differences (such as calculating the differences between temperature, humidity, light intensity, and gas concentration), when it is determined that the current external environmental characteristic values ​​are in a stable state (that is, substituting the current demand effect value and the pre-processed environmental difference into Y1=(s1+s2+s3)*X1+R1, Y2=(s1+s2+s3)*X2+R2, and Y3=(s1+s2+s3)*X3+R3 for testing to match three control schemes, and using the control scheme with the lowest energy consumption among the three predicted control schemes as the current predicted control scheme), or, when it is determined that the current external environmental characteristic values ​​are in a non-stable state... In a stable state (same as explained above, so no further explanation is needed here), the temperature and humidity control values, light intensity control values, and concentration control values ​​in the selected predicted control scheme are used to adjust the predicted parameter values ​​of the corresponding ventilation equipment (used to regulate air circulation in the farm, ensure air quality, provide sufficient oxygen for animals, and expel harmful gases): such as fans and ventilation fans; temperature control equipment (which will be activated to maintain a stable temperature in the farm when the ambient temperature is unsuitable): such as air conditioners, heaters, and coolers; and lighting systems (lighting equipment in the farm is used to provide illumination to meet the needs of animal growth and farm personnel operation).

[0089] An energy consumption control and regulation system for a smart farm, employing the energy consumption control and regulation method described above, includes a historical characteristic statistics unit one, a historical characteristic statistics unit two, a model building unit one, a model building unit two, and a current regulation scheme prediction unit, referencing... Figure 2The system acquires historical control records for intelligent energy consumption regulation during historical periods through the historical feature statistics unit one. If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and are subject to environmental regulation at the same control time point, then historical control scheme one is calculated. Similarly, if two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and are subject to environmental regulation at the same control time point, then historical control scheme two is calculated. Or, if any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental regulation at different control time points, then historical control scheme three is calculated. Based on historical control scheme one, historical control scheme two, and historical control scheme three, the model building unit one obtains the first training feature set, the second training feature set, and the third training feature set. If the external environmental characteristics of each historical period in the historical control record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set. If the external environmental characteristics of each historical period in the historical control record data are unstable, then from the historical change critical time point with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data, the first change control scheme implemented in the time period before the historical change critical time point, and the second change control scheme implemented in the time period after the historical change critical time point are statistically determined. Based on the first change control scheme and the second change control scheme, simulation prediction model two is trained. The current external environmental characteristic value of the current period is obtained through the current control scheme prediction unit, and it is determined whether the current external environmental characteristic value is in a stable state. Then, one of the simulation prediction models, simulation prediction model one and simulation prediction model two, is selected for testing to obtain the current prediction control scheme.

[0090] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for energy consumption control and regulation in a smart farm, characterized in that, Includes the following steps: Acquire historical control records of intelligent energy consumption control in historical periods. When three categories of data in the historical control records—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme one is calculated. When two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control records belong to the same historical period and are subject to environmental control at the same control time point, then historical control scheme two is calculated. Alternatively, when three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental control at different control time points, then historical control scheme three is calculated. Based on the historical control scheme 1, historical control scheme 2 and historical control scheme 3, a first training feature set, a second training feature set and a third training feature set are obtained. If the external environmental characteristics of each historical period in the historical control record data are stable, then a simulation prediction model 1 is trained based on the first training feature set, the second training feature set and the third training feature set. If the external environmental characteristics of each historical period in the historical control record data are unstable, then from the historical change critical time point with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data, the first change control scheme implemented in the time period before the historical change critical time point and the second change control scheme implemented in the time period after the historical change critical time point are statistically determined. Based on the first change control scheme and the second change control scheme, simulation prediction model two is trained. If the external environmental characteristics of each historical period in the historical control record data are unstable, then start from the critical time point of the historical change with the greatest degree of change in the external environmental characteristics of each historical period in the historical control record data. If one of the historical control schemes 1, 2, and 3 is implemented within the time period before the critical point of the historical change, then the first change control scheme is output. Furthermore, if a control scheme different from the first change control scheme is implemented among historical control scheme one, historical control scheme two, and historical control scheme three within the time period after the critical point of historical change, then the second change control scheme will be output. A fourth correlation coefficient is obtained by extracting the degree of mutual influence between the first change control scheme implemented in the period before the critical point of historical change within the same historical period and the second change control scheme implemented in the period after the critical point of historical change. Based on the first variable control scheme, the second variable control scheme, historical actual energy consumption values ​​and historical demand effect values, simulation prediction model two is trained. Obtain the current external environmental characteristic values ​​for the current period, determine whether the current external environmental characteristic values ​​are in a stable state, and select one of the simulation prediction models from simulation prediction model one and simulation prediction model two to test and obtain the current prediction and control scheme.

2. The energy consumption control and regulation method for a smart farm according to claim 1, characterized in that, Acquire historical control records of intelligent energy consumption regulation over a historical period. When three categories of data—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and were subject to environmental regulation at the same control time point, then the steps for calculating historical control scheme one are as follows: Acquire historical control records of intelligent energy consumption regulation in historical periods, and extract historical temperature and humidity control values, historical light intensity control values, and historical gas concentration control values ​​from the historical control records. If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and environmental control is performed at the same control time point, then historical test case one will be output. Based on the historical test case one, the first correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, historical light intensity control values ​​and historical concentration control values. The historical test scenario 1 and the first correlation coefficient are combined to form historical control scheme 1.

3. The energy consumption control and regulation method for a smart farm according to claim 2, characterized in that, When two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control records belong to the same historical period and were subject to environmental control at the same control time point, then historical control scheme two is calculated. Alternatively, when three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and were subject to environmental control at different control time points, then the steps for calculating historical control scheme three are as follows: If two categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to the same historical period and are subject to environmental control at the same control time point, then output historical test case two. Based on the second historical test case, the second correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, the historical light intensity control values, and the historical concentration control values. The second historical test scenario and the second correlation coefficient are combined to form the second historical control scheme. If any three categories of historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and environmental control is carried out at different control time points, then output historical test case three; Based on the aforementioned historical test case three, the third correlation coefficient is obtained by extracting the degree of mutual influence coefficient between the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values. The historical test scenario three and the third correlation coefficient are combined to form historical control scheme three.

4. The energy consumption control and regulation method for a smart farm according to claim 3, characterized in that, Based on the historical control schemes one, two, and three, a first training feature set, a second training feature set, and a third training feature set are obtained. If the external environmental characteristics of each historical period in the historical control record data are stable, then the steps for training simulation prediction model one based on the first, second, and third training feature sets are as follows: Extract the actual energy consumption values ​​of historical regulation after regulation in each historical period from the historical regulation record data, and extract the historical demand effect values ​​of each historical period before regulation in each historical period from the historical regulation record data. The historical regulation scheme 1, the historical actual energy consumption value of regulation, and the historical demand effect value are combined to form the first training feature set; The historical regulation scheme 2, the historical actual energy consumption value of regulation, and the historical demand effect value are combined to form the second training feature set; The historical regulation scheme three, the historical regulation actual energy consumption value and the historical demand effect value are combined to form the third training feature set; If the external environmental characteristics of each historical period in the historical regulation record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set.

5. The energy consumption control and regulation method for a smart farm according to claim 1, characterized in that, The steps of obtaining the current external environmental characteristic values ​​for the current period, determining whether the current external environmental characteristic values ​​are in a stable state, and selecting one of the simulation prediction models from simulation prediction model one and simulation prediction model two for testing to obtain the current prediction and control scheme are as follows: Obtain the current demand effect value and current external environmental characteristic value that require energy consumption regulation in the current period, and detect the demand environmental characteristic value that needs to be achieved in the current period; The difference between the current external environmental characteristic value and the required environmental characteristic value is calculated to obtain the preprocessed environmental difference value; When it is determined that the current external environmental characteristic values ​​are in a stable state, the current demand effect value and the pre-processed environmental difference value are input into the simulation prediction model one for testing to obtain the current prediction and control scheme. Alternatively, when it is determined that the current external environmental characteristic value is in an unstable state, the critical time point of the current change with the greatest degree of change in the external environmental characteristic is statistically identified. The current demand effect value, preprocessing environment difference, and current critical time point are then input into simulation prediction model two for testing to obtain the current prediction and control scheme.

6. An energy consumption control and regulation system for a smart farm, characterized in that, The system is used to implement the energy consumption control and regulation method for smart farms as described in any one of claims 1-5, including: Historical feature statistics unit one is used to acquire historical control records of intelligent energy consumption control in historical periods. When three categories of data in the historical control records—historical temperature and humidity control values, historical light intensity control values, and historical concentration control values—belong to the same historical period and environmental control is carried out at the same control time point, then historical control scheme one is statistically calculated. Historical feature statistics unit two is used to calculate historical control scheme two when two of the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​in the historical control record data belong to the same historical period and are subject to environmental control at the same control time point; or, when three of the historical temperature and humidity control values, historical light intensity control values, and historical concentration control values ​​belong to different historical periods and are subject to environmental control at different control time points, historical control scheme three is calculated. Model building unit one is used to obtain the first training feature set, the second training feature set, and the third training feature set based on historical control scheme one, historical control scheme two, and historical control scheme three. If the external environmental characteristics of each historical period in the historical control record data are stable, then simulation prediction model one is trained based on the first training feature set, the second training feature set, and the third training feature set. Model building unit two is used to calculate the first change control scheme in the period before the historical change critical time point and the second change control scheme in the period after the historical change critical time point if the external environmental characteristics of each historical period in the historical control record data are unstable. Then, the simulation prediction model two is trained based on the first change control scheme and the second change control scheme. The current control scheme prediction unit is used to obtain the current external environmental characteristic values ​​for the current period, determine whether the current external environmental characteristic values ​​are in a stable state, and select one of the simulation prediction models from simulation prediction model one and simulation prediction model two to test and obtain the current prediction control scheme.

Images