A remote control system and method for television sets based on the Internet of Things

Abstract

This invention discloses a remote control system and method for televisions based on the Internet of Things (IoT), relating to the field of remote control technology. The invention extracts the connection lines of all televisions, divides the connection lines into control units, collects device parameters for each television, and constructs a device profile in the cloud. A prediction model is trained using a neural network algorithm, and this model predicts the future power consumption of each control unit. A state percentage prediction model is generated using a neural network algorithm to predict the percentage of different states of the televisions within the control unit. The baseline power consumption of the control unit is calculated during peak hours, and control task indicators are established. During peak hours in each control unit, the televisions are classified into different device levels, and the adjustable capacity for each level is set. A comprehensive analysis is conducted to formulate a control instruction set. The control instructions are executed in the corresponding control unit, the control error is calculated, and the peak-shaving results are verified based on the judgment of the control error.

Description

Technical Field

[0001] This invention relates to the field of remote control technology, specifically to a remote control system and method for television sets based on the Internet of Things (IoT). Background Technology

[0002] With technological advancements and evolving consumer demands, the smart home appliance industry is gradually shifting from "single device control" to "ecological and intelligent management." Among these, the development of remote control technology for televisions, as the core entertainment terminal in the home, is increasingly fraught with contradictions between the needs of power grid energy management and the demands of remote control technology. On the one hand, the smart home appliance remote control industry has entered a mature development stage. Industry research shows that my country's smart home appliance remote control technology has evolved from early wired remote control to wireless remote control, and now to the "cloud-edge-device" collaborative control model based on the Internet of Things and cloud computing. Televisions, as one of the earliest home appliances to achieve remote control, have expanded their functions from basic "power on / off and channel switching" to "status monitoring and personalized mode adjustment." However, existing technologies largely focus on "individual user experience" and do not consider "the impact of clustered device control on the power grid load." On the other hand, the power grid faces the problem of a surge in load pressure during peak residential electricity consumption periods. With the increasing number of large-screen smart TVs, their rated power and average daily usage time have increased significantly. As a result, during peak electricity consumption periods such as 7:00 PM to 10:00 PM in the summer, the power consumption of TV clusters accounts for more than 30% of the total power consumption of households. When combined with the power consumption of equipment such as air conditioners and refrigerators, this can easily cause the load rate of the regional power grid transformer to exceed 85%, leading to problems such as voltage fluctuations and unstable power supply. There is an urgent need to achieve "peak shaving and valley filling" through remote control of TVs to balance the power grid load. Summary of the Invention

[0003] The purpose of this invention is to provide a remote control system and method for television sets based on the Internet of Things (IoT) to solve the problems raised in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A method for remotely controlling a television set based on the Internet of Things (IoT), the method comprising the following steps: S100: Extract all TV connection lines, divide control units for all TV connection lines, collect device parameters for each TV and build device profiles in the cloud. Furthermore, the specific steps for collecting device parameters for each television set and building a device profile in the cloud are as follows: S101. Extract the power grid transformer area connected to each television set, and group television sets connected to the same power grid transformer area into the same control unit. Finally, divide all television sets into control units for different areas. Each television set registers an account identity in the cloud. During registration, the cloud collects device data for each television set. The device data includes, but is not limited to, geographical location, device model, screen size, and rated power. The cloud constructs a device profile based on the device data of each television set. Specifically, a tag generation algorithm is used to tag each television set with different device data, and a visualization tool is used to display the tags of the television sets to obtain the device profile. In each control unit, all television sets are integrated into a device cluster. When the cloud controls the television sets, the device cluster within each control unit is the control target.

[0005] The control units are divided according to the power grid transformer area connected to the TV. This ensures that the power supply source of the equipment in each unit is consistent. Subsequent power consumption statistics and control command execution can focus on specific power supply areas, avoiding inefficiency or resource waste caused by cross-regional control. For example, when the load of a certain transformer area is too high, control can be applied only to the TVs in that area without affecting the normal power use of other areas.

[0006] By integrating all televisions within each control unit into a device cluster, the cloud uses the cluster as the control target, eliminating the need to operate each individual device individually. This significantly reduces the number of control commands issued, improves the response speed and execution efficiency of remote control, and is especially suitable for managing large-scale television clusters.

[0007] S200: Collect the total power consumption, time data and environmental data of each control unit in history, use a neural network algorithm to train a prediction model, and use the prediction model to predict the future power consumption of each control unit. Furthermore, the specific steps for predicting the future power consumption of each control unit using the predictive model are as follows: S201. Collect the total power consumption of each control unit per hour in history. When collecting based on time series, mark the features for different time characteristics. The time characteristics include weekdays, holidays and weekends. Mark the corresponding time characteristics in the time series corresponding to the total power consumption. S202. Using a neural network algorithm, train the collected time-series-based total power consumption to construct a prediction model, the formula of which is: ; In the formula, L pred (t) represents the predicted power consumption at time t, Seq2Seq represents the sequence-to-sequence model in the neural network algorithm, and L hist F represents the total power consumption of the control unit in the collected history. time(t) represents the time characteristic at time t corresponding to the total power consumption of the collected control unit; S203. Input the time series of the next 24 hours and the corresponding time features into the prediction model, mark the collected time features, input them into the prediction model, and output the predicted power consumption of the control unit for the next 24 hours.

[0008] Historical total power consumption is collected by marking time features such as weekdays, holidays, and weekends, enabling the model to fully learn power consumption patterns under different time scenarios. For example, families typically watch more television on weekends, resulting in higher power consumption; the model can use this feature to more accurately predict weekend power consumption.

[0009] By predicting the power consumption of each control unit in the next 24 hours, staff can know the trend of power consumption changes in advance, avoiding the hasty formulation of control plans after the peak power consumption arrives, thus gaining sufficient time for control preparation and improving the initiative and foresight of control.

[0010] S300: Classify the TV state into three states: power on, standby, and power off; collect historical time-series data on the proportion of each of the three states of the TV in the control unit; use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. Furthermore, the specific steps for predicting the proportion of different states of television sets within the future control unit are as follows: S301. Collect historical data on the proportion and time characteristics of the number of each type of television set in the three states within the control unit, based on historical time series data. Use the time characteristics to label the time series data. Use a neural network algorithm to train the collected historical proportions to obtain a state proportion prediction model. S302. Input the time series of the next 24 hours and the corresponding time features into the state proportion prediction model. The state proportion prediction model outputs the proportion of the three types of TV states in the next 24 hours.

[0011] By categorizing television settings into three types—power on, standby, and power off—and collecting historical state percentages, combined with time-related data to train a model, we can clearly understand the patterns in the number of televisions in each state at different times. For example, during weekdays, most household televisions are in power off or standby mode, and the percentage data will exhibit specific characteristics.

[0012] Predicting the percentage of different states in the next 24 hours can be directly used to calculate subsequent regulatory potential. If accurate state percentage data is lacking, the calculation of regulatory potential will have a large error, affecting the rationality of the formulation of regulatory task indicators.

[0013] S400: Preset power consumption threshold, extract peak periods from predicted power consumption, use peak periods to calculate the baseline power consumption of the control unit, use the baseline power consumption to calculate the control potential of each control unit, calculate the control demand during peak periods, and formulate control task indicators. Furthermore, the specific steps for formulating regulatory targets are as follows: S401. Staff preset a power consumption threshold and use the power consumption threshold to judge the predicted power consumption of the control unit for the next 24 hours. One hour is a unit of time. When the predicted power consumption per unit of time in the next 24 hours is greater than the power consumption threshold, it is judged as high power consumption. After judging the predicted power consumption for 24 hours, when the unit time of high power consumption is judged to be continuous, all continuous high power consumption unit times are defined as peak periods. S402. Extract the predicted power consumption during off-peak hours in the next 24 hours, and calculate the average predicted power consumption during off-peak hours as the baseline power consumption; calculate the control demand during peak hours within the control unit, using the following formula: ; In the formula, △L total (t) represents the control demand within the control unit at time t during peak period, L peak (t) represents the predicted power consumption during peak period t, L base (t) represents the reference power consumption of the control unit; S403. During peak hours, the number of each of the three predicted television states is calculated by multiplying the predicted percentage of television sets in each state by the total number of television sets. The control potential within the control unit is then calculated using the following formula: ; In the formula, P potential (t) represents the control potential within the control unit t during peak period, N on (t) represents the predicted number of TVs powered on during peak period t, α represents the power consumption adjustment coefficient during power-on, and P on avg N represents the average power consumption of a television when it is powered on. standby (t) represents the predicted number of TVs in standby mode during peak period t, β represents the standby power consumption adjustment coefficient, and P standby This represents the average power consumption of the television in standby mode. S404. Develop control task indicators using the control demand and control potential within the control unit, using the following formula: ; In the formula, △L target (t) represents the control task indicator at peak time t, and λ represents the safety factor, which is set by the staff.

[0014] Using the average predicted power consumption during off-peak hours as the benchmark power consumption, this value can reflect the power consumption level of the control unit under normal and low load conditions, providing a reasonable reference for calculating control requirements, making the calculation of control requirements more in line with the actual situation, and avoiding unreasonable control targets due to excessively high or low benchmark values.

[0015] The potential for regulation is calculated by combining formulas with state percentages and device power consumption parameters. Regulation requirements are then calculated by combining predicted power consumption with baseline power consumption, ensuring that the quantitative calculation of regulation potential and requirements is scientific and accurate. Based on this, regulation task indicators are formulated using the minimum of the two values ​​combined with a safety factor. This ensures that the power reduction requirements are met while avoiding excessive regulation that could impact normal user operation.

[0016] S500: During peak hours of each control unit, the television is classified into different equipment levels, the adjustable capacity of each equipment level is set, and a set of control instructions is formulated through comprehensive analysis. Furthermore, the specific steps for comprehensively analyzing and formulating the control instruction set are as follows: S501. When the real-time time is in the predicted peak period, collect the user operation frequency of all TVs in the power-on state, set the frequency threshold, and regard TVs in the power-on state with a frequency greater than the frequency threshold as high-value power-on devices and TVs in the power-on state with a frequency less than the frequency threshold as low-value power-on devices; set high-value power-on devices as grade A, low-value power-on devices as grade B, and standby TVs as grade C. S502. The formula for calculating the adjustable capacity of Class B and Class C televisions is as follows: ; In the formula, Cap B P indicates the adjustable capacity of a Class B television. current This indicates the real-time power consumption of TV B; w represents the adjustable ratio; Cap C =P standby , among which, Cap C This indicates the adjustable capacity of a Class C television. By calculating the adjustable capacity of Class B and Class C equipment using formulas, staff can clearly understand the power consumption adjustment range that each type of equipment can achieve, providing data support for control sequencing and instruction formulation, and avoiding blind control that leads to excessive or insufficient power consumption adjustment.

[0017] S503. Based on the calculated control task indicators of the control unit during peak hours, formulate a control instruction set, specifically: during peak hours, control priority level C is higher than level B. After the control of level C televisions is completed, control of level B televisions is then performed. When controlling level B televisions, the adjustable capacity of level B televisions is used to sort the control order, and control is performed in descending order. During peak hours, the power consumption of televisions is adjusted sequentially according to the above control rules until the control task indicators are completed. S504. The control strategy for Class B televisions is as follows: if no user operates the screen to prompt entry into audio mode, the audio mode will turn off the screen and only play audio; the control strategy for Class C televisions is as follows: the television displays: "To cooperate with the power grid's energy saving, the device will start up after a delay of [ΔT] seconds. Thank you for your support," and a random delay time ΔT is set. The standby television will then start up again after time ΔT.

[0018] The C-level control priority is set higher than the B-level, and the B-level controls are adjusted from the largest to the smallest adjustable capacity. This ensures that the standby devices that have the least impact on users are controlled first, followed by the low-value powered-on devices. At the same time, devices with high adjustment potential are prioritized to improve the control effect. Specific strategies such as audio mode and delayed start are formulated for B and C-level devices respectively to make the control commands more executable.

[0019] S600 executes the control instructions in the corresponding control unit, collects the expected control amount and actual regional power consumption of each device during control, calculates the control error, and verifies the peak shaving results based on the judgment of the control error.

[0020] Furthermore, the specific steps for optimizing the control instruction set based on the judgment of control error are as follows: S601. When the system issues control commands to each television set according to the control command set, it extracts the control amount from the corresponding control command for each television set and sums them to obtain the expected control amount of the control unit. It then uses a meter to collect the actual total power consumption within the control unit and calculates the control error using the following formula: ; In the formula, Dev(t) represents the control error, and L act (t) represents the actual total power consumption, ΔL exp Indicates the expected amount of regulation; S602. Set an error threshold. When the control error is greater than the error threshold, the peak shaving is considered successful; when the control error is less than the error threshold, the peak shaving is considered unsuccessful.

[0021] A threshold error is set, and the success of peak shaving is determined based on the relationship between the control error and the threshold. If peak shaving fails, staff can analyze the cause in a timely manner and optimize and adjust the subsequent control scheme to avoid the problem persisting and the power consumption peak becoming unmanageable.

[0022] A remote control system for televisions based on the Internet of Things (IoT) includes a data acquisition module, a profile construction module, a prediction module, a control module, and a verification module. The data acquisition module is used to extract the connection lines of all TVs, divide the control units for the connection lines of all TVs, and collect the equipment parameters of each TV. The profile building module is used to build a device profile in the cloud based on the device data of each TV. The prediction module is used to train a power consumption prediction model using a neural network algorithm, and to use the prediction model to predict the predicted power consumption of each control unit in the future; and to use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. The control module is used to calculate the control potential of each control unit using the reference power consumption, calculate the control demand during peak hours, formulate control task indicators, classify televisions into different equipment levels, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set. The verification module is used to collect the expected control amount and actual regional power consumption of each device during control, calculate the control error, and verify the peak-shaving results based on the judgment of the control error.

[0023] The prediction module includes a power consumption prediction unit and a state percentage prediction unit; The power consumption prediction unit is used to collect the total power consumption, time data and environmental data of each control unit in history, train a prediction model using a neural network algorithm, and use the prediction model to predict the future power consumption of each control unit. The state proportion prediction unit is used to classify the television state into three states: power on, standby, and power off. It collects the proportion of each of the three states of each type of television in the control unit based on historical time series data, and uses a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the television in the control unit in the future.

[0024] The control module includes a control task calculation unit and a control instruction unit; The control task calculation unit is used to preset power consumption thresholds, extract peak periods from predicted power consumption, calculate the baseline power consumption of the control unit using peak periods, calculate the control potential of each control unit using the baseline power consumption, calculate the control demand during peak periods, and formulate control task indicators. The control instruction unit is used to classify the television set into different equipment levels during peak hours of each control unit, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set.

[0025] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention is based on data in every step, from equipment data acquisition and model training and prediction to the formulation of control indicators, command execution and verification. Through quantitative calculation and scientific models, it avoids the empiricism and blindness in traditional control, making the whole control process accurate and efficient, and can effectively achieve the power consumption peak reduction target of TV clusters.

[0026] 2. In the control design, by classifying equipment, setting priorities, and implementing targeted control strategies, the user experience is fully considered. This avoids excessive interference with normal user operation in pursuit of energy saving, achieving a balance between energy-saving goals and user experience, and making it more acceptable to users.

[0027] 3. This invention effectively reduces total power consumption during peak hours by remotely controlling the television cluster, thereby reducing the load pressure on the power grid during peak hours and avoiding problems such as power grid fluctuations and unstable power supply caused by excessive load. This provides a guarantee for the safe and stable operation of the power grid and is also in line with the overall development trend of energy conservation and emission reduction. Attached Figure Description

[0028] Figure 1 This is a module distribution diagram of a remote control system for a television set based on the Internet of Things (IoT) according to the present invention. Figure 2 This is a schematic diagram illustrating the steps of a remote control method for a television set based on the Internet of Things (IoT) according to the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Example: Figures 1-2 As shown, the present invention provides a technical solution. A method for remotely controlling a television set based on the Internet of Things (IoT), the method comprising the following steps: S100: Extract all TV connection lines, divide control units for all TV connection lines, collect device parameters for each TV and build device profiles in the cloud. The specific steps for collecting device parameters for each TV and building a device profile in the cloud are as follows: S101. Extract the power grid transformer area connected to each television set, and group television sets connected to the same power grid transformer area into the same control unit. Finally, divide all television sets into control units for different areas. Each television set registers an account identity in the cloud. During registration, the cloud collects device data for each television set. The device data includes, but is not limited to, geographical location, device model, screen size, and rated power. The cloud constructs a device profile based on the device data of each television set. Specifically, a tag generation algorithm is used to tag each television set with different device data, and a visualization tool is used to display the tags of the television sets to obtain the device profile. In each control unit, all television sets are integrated into a device cluster. When the cloud controls the television sets, the device cluster within each control unit is the control target.

[0031] The control units are divided according to the power grid transformer area connected to the TV. This ensures that the power supply source of the equipment in each unit is consistent. Subsequent power consumption statistics and control command execution can focus on specific power supply areas, avoiding inefficiency or resource waste caused by cross-regional control. For example, when the load of a certain transformer area is too high, control can be applied only to the TVs in that area without affecting the normal power use of other areas.

[0032] By integrating all televisions within each control unit into a device cluster, the cloud uses the cluster as the control target, eliminating the need to operate each individual device individually. This significantly reduces the number of control commands issued, improves the response speed and execution efficiency of remote control, and is especially suitable for managing large-scale television clusters.

[0033] S200: Collect the total power consumption, time data and environmental data of each control unit in history, use a neural network algorithm to train a prediction model, and use the prediction model to predict the future power consumption of each control unit. The specific steps for predicting the future power consumption of each control unit using a predictive model are as follows: S201. Collect the total power consumption of each control unit per hour in history. When collecting based on time series, mark the features for different time characteristics. The time characteristics include weekdays, holidays and weekends. Mark the corresponding time characteristics in the time series corresponding to the total power consumption. S202. Using a neural network algorithm, train the collected time-series-based total power consumption to construct a prediction model, the formula of which is: ; In the formula, L pred (t) represents the predicted power consumption at time t, Seq2Seq represents the sequence-to-sequence model in the neural network algorithm, and L hist F represents the total power consumption of the control unit in the collected history. time(t) represents the time characteristic at time t corresponding to the total power consumption of the collected control unit; S203. Input the time series of the next 24 hours and the corresponding time features into the prediction model, mark the collected time features, input them into the prediction model, and output the predicted power consumption of the control unit for the next 24 hours.

[0034] Historical total power consumption is collected by marking time features such as weekdays, holidays, and weekends, enabling the model to fully learn power consumption patterns under different time scenarios. For example, families typically watch more television on weekends, resulting in higher power consumption; the model can use this feature to more accurately predict weekend power consumption.

[0035] By predicting the power consumption of each control unit in the next 24 hours, staff can know the trend of power consumption changes in advance, avoiding the hasty formulation of control plans after the peak power consumption arrives, thus gaining sufficient time for control preparation and improving the initiative and foresight of control.

[0036] S300: Classify the TV state into three states: power on, standby, and power off; collect historical time-series data on the proportion of each of the three states of the TV in the control unit; use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. The specific steps for predicting the percentage of different states of televisions within the future control unit are as follows: S301. Collect historical data on the proportion and time characteristics of the number of each type of television set in the three states within the control unit, based on historical time series data. Use the time characteristics to label the time series data. Use a neural network algorithm to train the collected historical proportions to obtain a state proportion prediction model. S302. Input the time series of the next 24 hours and the corresponding time features into the state proportion prediction model. The state proportion prediction model outputs the proportion of the three types of TV states in the next 24 hours.

[0037] By categorizing television settings into three types—power on, standby, and power off—and collecting historical state percentages, combined with time-related data to train a model, we can clearly understand the patterns in the number of televisions in each state at different times. For example, during weekdays, most household televisions are in power off or standby mode, and the percentage data will exhibit specific characteristics.

[0038] Predicting the percentage of different states in the next 24 hours can be directly used to calculate subsequent regulatory potential. If accurate state percentage data is lacking, the calculation of regulatory potential will have a large error, affecting the rationality of the formulation of regulatory task indicators.

[0039] S400: Preset power consumption threshold, extract peak periods from predicted power consumption, use peak periods to calculate the baseline power consumption of the control unit, use the baseline power consumption to calculate the control potential of each control unit, calculate the control demand during peak periods, and formulate control task indicators. The specific steps for formulating regulatory targets are as follows: S401. Staff preset a power consumption threshold and use the power consumption threshold to judge the predicted power consumption of the control unit for the next 24 hours. One hour is a unit of time. When the predicted power consumption per unit of time in the next 24 hours is greater than the power consumption threshold, it is judged as high power consumption. After judging the predicted power consumption for 24 hours, when the unit time of high power consumption is judged to be continuous, all continuous high power consumption unit times are defined as peak periods. S402. Extract the predicted power consumption during off-peak hours in the next 24 hours, and calculate the average predicted power consumption during off-peak hours as the baseline power consumption; calculate the control demand during peak hours within the control unit, using the following formula: ; In the formula, △L total (t) represents the control demand within the control unit at time t during peak period, L peak (t) represents the predicted power consumption during peak period t, L base (t) represents the reference power consumption of the control unit; S403. During peak hours, the number of each of the three predicted television states is calculated by multiplying the predicted percentage of television sets in each state by the total number of television sets. The control potential within the control unit is then calculated using the following formula: ; In the formula, P potential (t) represents the control potential within the control unit t during peak period, N on (t) represents the predicted number of TVs powered on during peak period t, α represents the power consumption adjustment coefficient during power-on, and P on avg N represents the average power consumption of a television when it is powered on. standby (t) represents the predicted number of TVs in standby mode during peak period t, β represents the standby power consumption adjustment coefficient, and P standby This represents the average power consumption of the television in standby mode. S403. Develop control task indicators using the control demand and control potential within the control unit, using the following formula: ; In the formula, △L target (t) represents the control task indicator at peak time t, and λ represents the safety factor, which is set by the staff.

[0040] Using the average predicted power consumption during off-peak hours as the benchmark power consumption, this value can reflect the power consumption level of the control unit under normal and low load conditions, providing a reasonable reference for calculating control requirements, making the calculation of control requirements more in line with the actual situation, and avoiding unreasonable control targets due to excessively high or low benchmark values.

[0041] The potential for regulation is calculated by combining formulas with state percentages and device power consumption parameters. Regulation requirements are then calculated by combining predicted power consumption with baseline power consumption, ensuring that the quantitative calculation of regulation potential and requirements is scientific and accurate. Based on this, regulation task indicators are formulated using the minimum of the two values ​​combined with a safety factor. This ensures that the power reduction requirements are met while avoiding excessive regulation that could impact normal user operation.

[0042] S500: During peak hours of each control unit, the television is classified into different equipment levels, the adjustable capacity of each equipment level is set, and a set of control instructions is formulated through comprehensive analysis. The specific steps for formulating a comprehensive set of control instructions are as follows: S501. When the real-time time is in the predicted peak period, collect the user operation frequency of all TVs in the power-on state, set the frequency threshold, and regard TVs in the power-on state with a frequency greater than the frequency threshold as high-value power-on devices and TVs in the power-on state with a frequency less than the frequency threshold as low-value power-on devices; set high-value power-on devices as grade A, low-value power-on devices as grade B, and standby TVs as grade C. S502. The formula for calculating the adjustable capacity of Class B and Class C televisions is as follows: ; In the formula, Cap B P indicates the adjustable capacity of a Class B television. current This indicates the real-time power consumption of TV B; w represents the adjustable ratio; Cap C =P standby , among which, Cap C This indicates the adjustable capacity of a Class C television. By calculating the adjustable capacity of Class B and Class C equipment using formulas, staff can clearly understand the power consumption adjustment range that each type of equipment can achieve, providing data support for control sequencing and instruction formulation, and avoiding blind control that leads to excessive or insufficient power consumption adjustment.

[0043] S503. Based on the calculated control task indicators of the control unit during peak hours, formulate a control instruction set, specifically: during peak hours, control priority level C is higher than level B. After the control of level C televisions is completed, control of level B televisions is then performed. When controlling level B televisions, the adjustable capacity of level B televisions is used to sort the control order, and control is performed in descending order. During peak hours, the power consumption of televisions is adjusted sequentially according to the above control rules until the control task indicators are completed. S504. The control strategy for Class B televisions is as follows: if no user operates the screen to prompt entry into audio mode, the audio mode will turn off the screen and only play audio; the control strategy for Class C televisions is as follows: the television displays: "To cooperate with the power grid's energy saving, the device will start up after a delay of [ΔT] seconds. Thank you for your support," and a random delay time ΔT is set. The standby television will then start up again after time ΔT.

[0044] The C-level control priority is set higher than the B-level, and the B-level controls are adjusted from the largest to the smallest adjustable capacity. This ensures that the standby devices that have the least impact on users are controlled first, followed by the low-value powered-on devices. At the same time, devices with high adjustment potential are prioritized to improve the control effect. Specific strategies such as audio mode and delayed start are formulated for B and C-level devices respectively to make the control commands more executable.

[0045] S600 executes the control instructions in the corresponding control unit, collects the expected control amount and actual regional power consumption of each device during control, calculates the control error, and verifies the peak shaving results based on the judgment of the control error.

[0046] The specific steps for optimizing the control instruction set based on the judgment of control error are as follows: S601. When the system issues control commands to each television set according to the control command set, it extracts the control amount from the corresponding control command for each television set and sums them to obtain the expected control amount of the control unit. It then uses a meter to collect the actual total power consumption within the control unit and calculates the control error using the following formula: ; In the formula, Dev(t) represents the control error, and L act (t) represents the actual total power consumption, ΔL exp Indicates the expected amount of regulation; S602. Set an error threshold. When the control error is greater than the error threshold, the peak shaving is considered successful; when the control error is less than the error threshold, the peak shaving is considered unsuccessful.

[0047] A threshold error is set, and the success of peak shaving is determined based on the relationship between the control error and the threshold. If peak shaving fails, staff can analyze the cause in a timely manner and optimize and adjust the subsequent control scheme to avoid the problem persisting and the power consumption peak becoming unmanageable.

[0048] A remote control system for televisions based on the Internet of Things (IoT) includes a data acquisition module, a profile construction module, a prediction module, a control module, and a verification module. The data acquisition module is used to extract the connection lines of all TVs, divide the control units for the connection lines of all TVs, and collect the equipment parameters of each TV. The profile building module is used to build a device profile in the cloud based on the device data of each TV. The prediction module is used to train a power consumption prediction model using a neural network algorithm, and to use the prediction model to predict the predicted power consumption of each control unit in the future; and to use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. The control module is used to calculate the control potential of each control unit using the reference power consumption, calculate the control demand during peak hours, formulate control task indicators, classify televisions into different equipment levels, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set. The verification module is used to collect the expected control amount and actual regional power consumption of each device during control, calculate the control error, and verify the peak-shaving results based on the judgment of the control error.

[0049] The prediction module includes a power consumption prediction unit and a state percentage prediction unit; The power consumption prediction unit is used to collect the total power consumption, time data and environmental data of each control unit in history, train a prediction model using a neural network algorithm, and use the prediction model to predict the future power consumption of each control unit. The state proportion prediction unit is used to classify the television state into three states: power on, standby, and power off. It collects the proportion of each of the three states of each type of television in the control unit based on historical time series data, and uses a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the television in the control unit in the future.

[0050] The control module includes a control task calculation unit and a control instruction unit; The control task calculation unit is used to preset power consumption thresholds, extract peak periods from predicted power consumption, calculate the baseline power consumption of the control unit using peak periods, calculate the control potential of each control unit using the baseline power consumption, calculate the control demand during peak periods, and formulate control task indicators. The control instruction unit is used to classify the television set into different equipment levels during peak hours of each control unit, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set.

[0051] Example: Install smart meters in each transformer substation area of ​​the community (to collect the total power consumption of the substation area), and add IoT modules to the residents' TVs (supporting WiFi / 4G connection, collecting device on / off status, real-time power consumption, and user operation frequency). All data is transmitted to the cloud platform through the IoT gateway; the cloud registers and builds profiles for the TVs; the community is divided into control units A1-A10. Input the 15-day historical total power consumption data (including time feature markers) of unit A3 into the Seq2Seq model to train the power consumption prediction model; input the 24-hour time series of July 20 (Sunday, not a public holiday) and the "weekend" time features, and the model outputs the predicted power consumption: power consumption is 12-18kW from 7:00 to 18:00, power consumption rises to 25-32kW from 19:00 to 22:00, and falls back to below 10kW after 23:00.

[0052] The "TV status percentage - time feature" data of 15 days in Unit A3 was input into the neural network model to train the status percentage prediction model; the 24-hour time series and "weekend" feature of July 20 were input, and the model output: 60% of TVs were on (approximately 492 units) from 19:00 to 22:00, 25% were in standby (approximately 205 units), and 15% were off (approximately 123 units); the percentage of TVs on during other times was ≤30%.

[0053] The preset power consumption threshold is 22kW. Compared with the predicted power consumption on July 20th, it is determined that 19:00-22:00 is a continuous high power consumption period (4 hours), i.e., the peak period; the baseline power consumption L is calculated. base =14kW; Calculate the control demand at each time from 19:00 to 22:00: 19:00 (25-14=11kW), 20:00 (30-14=16kW), 21:00 (32-14=18kW), 22:00 (28-14=14kW); the control potential from 19:00 to 22:00 is approximately 19kW; calculate the task indicators at each time: 19:00 (min(11,19×0.9)=11kW), 20:00 (min(16,17.1)=16kW), 21:00 (min(18,17.1)=17.1kW), 22:00 (min(14,17.1)=14kW).

[0054] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for remotely controlling a television set based on the Internet of Things, characterized in that: The method includes the following steps: S100: Extract all TV connection lines, divide control units for all TV connection lines, collect device parameters for each TV and build device profiles in the cloud. S200: Collect the total power consumption, time data and environmental data of each control unit in history, use a neural network algorithm to train a prediction model, and use the prediction model to predict the future power consumption of each control unit. S300: Classify the TV state into three states: power on, standby, and power off; collect historical time-series data on the proportion of each of the three states of the TV in the control unit; use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. S400: Preset power consumption threshold, extract peak periods from predicted power consumption, use peak periods to calculate the baseline power consumption of the control unit, use the baseline power consumption to calculate the control potential of each control unit, calculate the control demand during peak periods, and formulate control task indicators. S500: During peak hours of each control unit, the television is classified into different equipment levels, the adjustable capacity of each equipment level is set, and a set of control instructions is formulated through comprehensive analysis. S600 executes the control instructions in the corresponding control unit, collects the expected control amount and actual regional power consumption of each device during control, calculates the control error, and verifies the peak shaving results based on the judgment of the control error.

2. The method for remotely controlling a television set based on the Internet of Things according to claim 1, characterized in that: The specific steps in S100 for collecting device parameters for each television set and building a device profile in the cloud are as follows: S101. Extract the power grid transformer area connected to each TV set, and classify TV sets connected to the same power grid transformer area into the same control unit. Finally, classify all TV sets into control units for different areas. Each TV set registers an account identity in the cloud. During registration, the cloud collects the device data of each TV set. The device data includes, but is not limited to, geographical location, device model, screen size, and rated power. The cloud constructs a device profile based on the device data of each TV set. Specifically, a tag generation algorithm is used to tag each TV set with different device data, and a visualization tool is used to display the tags of the TV set to obtain the device profile. In each control unit, all televisions are integrated into a device cluster. When the cloud controls the televisions, the device cluster within each control unit is the control target.

3. The method for remotely controlling a television set based on the Internet of Things according to claim 2, characterized in that: The specific steps in S200 for predicting the future power consumption of each control unit using a prediction model are as follows: S201. Collect the total power consumption of each control unit per hour in history. When collecting based on time series, mark the features for different time characteristics. The time characteristics include weekdays, holidays and weekends. Mark the corresponding time characteristics in the time series corresponding to the total power consumption. S202. Using a neural network algorithm, train the collected time-series-based total power consumption to construct a prediction model, the formula of which is: ； In the formula, L pred (t) represents the predicted power consumption at time t, Seq2Seq represents a sequence-to-sequence model in a neural network algorithm, L hist represents the total power consumption of the regulation unit in the collected history, F time (t) represents the time characteristics of the total power consumption of the regulation unit corresponding to time t. S203. Input the time series of the next 24 hours and the corresponding time features into the prediction model, mark the collected time features, input them into the prediction model, and output the predicted power consumption of the control unit for the next 24 hours.

4. The method for remotely controlling a television set based on the Internet of Things according to claim 3, characterized in that: The specific steps in S300 for predicting the proportion of different states of the television set within the future control unit are as follows: S301. Collect historical data on the proportion and time characteristics of the number of each type of television set in the three states within the control unit, based on historical time series data. Use the time characteristics to label the time series data. Use a neural network algorithm to train the collected historical proportions to obtain a state proportion prediction model. S302. Input the time series of the next 24 hours and the corresponding time features into the state proportion prediction model. The state proportion prediction model outputs the proportion of the three types of TV states in the next 24 hours.

5. The method for remotely controlling a television set based on the Internet of Things according to claim 4, characterized in that: The specific steps for setting control task indicators in S400 are as follows: S401. Staff preset a power consumption threshold and use the power consumption threshold to judge the predicted power consumption of the control unit for the next 24 hours. One hour is a unit of time. When the predicted power consumption per unit of time in the next 24 hours is greater than the power consumption threshold, it is judged as high power consumption. After judging the predicted power consumption for 24 hours, when the unit time of high power consumption is judged to be continuous, all continuous high power consumption unit times are defined as peak periods. S402. Extract the predicted power consumption during off-peak hours in the next 24 hours, and calculate the average predicted power consumption during off-peak hours as the baseline power consumption. The formula for calculating the control demand during peak hours within the control unit is as follows: ； In the formula, ΔL total (t) represents the control demand in the control unit at peak time t, L peak (t) represents the predicted power consumption at peak time t, L base (t) represents the reference power consumption of the control unit; S403. During peak hours, the number of each of the three predicted television states is calculated by multiplying the predicted percentage of television sets in each state by the total number of television sets. The control potential within the control unit is then calculated using the following formula: ； In the formula, P potential (t) represents the control potential in the peak period t control unit, N on (t) represents the predicted number of on state TV in peak period t, α represents the on state power consumption adjustment coefficient, P on avg represents the average power consumption of the on state TV; N standby (t) represents the predicted number of standby state TV in peak period t, β represents the standby state power consumption adjustment coefficient, P standby represents the average power consumption of the standby state TV; S404. Develop control task indicators using the control demand and control potential within the control unit, using the following formula: ； In the formula, ΔL target (t) represents the control task index at peak time t, and λ represents the safety factor, which is set by the staff.

6. The method for remotely controlling a television set based on the Internet of Things according to claim 5, characterized in that: The specific steps for formulating the control instruction set through comprehensive analysis in the S500 are as follows: S501. When the real-time time is in the predicted peak period, collect the user operation frequency of all TVs in the power-on state, set the frequency threshold, and regard TVs in the power-on state with a frequency greater than the frequency threshold as high-value power-on devices and TVs in the power-on state with a frequency less than the frequency threshold as low-value power-on devices; set high-value power-on devices as grade A, low-value power-on devices as grade B, and standby TVs as grade C. S502. The formula for calculating the adjustable capacity of Class B and Class C televisions is as follows: ； In the formula, Cap B represents the adjustable capacity of a B-class TV set, P current represents the real-time power consumption of a B-class TV set, and w represents the adjustable proportion. Cap C = P standby , where Cap C represents the adjustable capacity of a C-class TV set; S503. Based on the calculated control task indicators of the control unit during peak hours, formulate a control instruction set, specifically: during peak hours, control priority level C is higher than level B. After the control of level C televisions is completed, control of level B televisions is then performed. When controlling level B televisions, the adjustable capacity of level B televisions is used to sort the control order, and control is performed in descending order. During peak hours, the power consumption of televisions is adjusted sequentially according to the above control rules until the control task indicators are completed. S504. The control strategy for Class B televisions is as follows: if no user operates the screen to prompt entry into audio mode, the audio mode will turn off the screen and only play audio; the control strategy for Class C televisions is as follows: the television displays: "To cooperate with the power grid's energy saving, the device will start up after a delay of [ΔT] seconds. Thank you for your support," and a random delay time ΔT is set. The standby television will then start up again after time ΔT.

7. The method for remotely controlling a television set based on the Internet of Things according to claim 6, characterized in that: The specific steps in S600 for optimizing the control instruction set based on the judgment of control error are as follows: S601. When the system issues control commands to each television set according to the control command set, it extracts the control amount from the corresponding control command for each television set and sums them to obtain the expected control amount of the control unit. It then uses a meter to collect the actual total power consumption within the control unit and calculates the control error using the following formula: ； In the formula, Dev(t) represents a control error, L act (t) represents an actual total power consumption, ΔL exp represents an expected control amount; S602. Set an error threshold. When the control error is greater than the error threshold, the peak shaving is considered successful; when the control error is less than the error threshold, the peak shaving is considered unsuccessful.

8. A remote control system for televisions based on the Internet of Things, characterized in that: The remote control system for television sets includes a data acquisition module, a profile building module, a prediction module, a control module, and a verification module. The data acquisition module is used to extract the connection lines of all TVs, divide the control units for the connection lines of all TVs, and collect the equipment parameters of each TV. The profile building module is used to build a device profile in the cloud based on the device data of each TV. The prediction module is used to train a power consumption prediction model using a neural network algorithm, and to use the prediction model to predict the predicted power consumption of each control unit in the future; and to use a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the TV in the control unit in the future. The control module is used to calculate the control potential of each control unit using the reference power consumption, calculate the control demand during peak hours, formulate control task indicators, classify televisions into different equipment levels, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set. The verification module is used to collect the expected control amount and actual regional power consumption of each device during control, calculate the control error, and verify the peak-shaving results based on the judgment of the control error.

9. A remote control system for a television set based on the Internet of Things according to claim 8, characterized in that: The prediction module includes a power consumption prediction unit and a state percentage prediction unit; The power consumption prediction unit is used to collect the total power consumption, time data and environmental data of each control unit in history, train a prediction model using a neural network algorithm, and use the prediction model to predict the future power consumption of each control unit. The state proportion prediction unit is used to classify the television state into three states: power on, standby, and power off. It collects the proportion of each of the three states of each type of television in the control unit based on historical time series data, and uses a neural network algorithm to generate a state proportion prediction model to predict the proportion of different states of the television in the control unit in the future.

10. A remote control system for a television set based on the Internet of Things according to claim 8, characterized in that: The control module includes a control task calculation unit and a control instruction unit; The control task calculation unit is used to preset power consumption thresholds, extract peak periods from predicted power consumption, calculate the baseline power consumption of the control unit using peak periods, calculate the control potential of each control unit using the baseline power consumption, calculate the control demand during peak periods, and formulate control task indicators. The control instruction unit is used to classify the television set into different equipment levels during peak hours of each control unit, set the adjustable capacity for each equipment level, and comprehensively analyze and formulate a control instruction set.

Images