SEHWS energy prediction method based on double-index moving average and SEH wireless sensor

Abstract

The invention provides an SEHWS energy prediction method based on double-index moving average and an SEH wireless sensor, and belongs to the technical field of energy prediction of solar energy collection wireless sensors. The method comprises the following steps: collecting and processing historical energy collection data, and constructing a historical D-day energy matrix E; calculating parameters mu D (d, n), B (d, n), omega and gamma in the model; and substituting the four parameters into the DEMA energy prediction model based on historical data, and carrying out weighted calculation on the collected energy predicted in the previous time slot and the average value of the actually collected energy information of the current time slot n in the past several days to obtain a predicted value E (-) (d, n) of the collected energy in the n time slot. The SEH wireless sensor is composed of a microcontroller, a data acquisition module, a solar energy collection module, a memory, a wireless communication module and a power management module.

Description

Technical Field

[0001] The present invention relates to the technical field of solar energy collection wireless sensor energy prediction, and in particular to a SEHWS energy prediction method based on double exponential moving average and a SEH wireless sensor. Background Art

[0002] With the rapid development of Internet of Things technology, more and more devices and systems need to be interconnected, resulting in a huge demand for wireless sensors. Wireless sensor networks are widely used in environmental monitoring, human health detection, public safety and other fields. They are composed of multiple wireless sensor nodes, which are distributed in the monitoring area. They sense the physical or chemical quantities in the environment, perform preliminary data processing, and transmit data through wireless communication protocols. The data is transmitted to the aggregation node in a single hop or multi-hop manner. The aggregation node then transmits the data to the central server or cloud platform for centralized processing and storage. Throughout the process, wireless sensors usually rely on batteries for power supply, but the capacity of traditional batteries is limited, resulting in a limited life of their nodes. When the battery is exhausted, the wireless sensor node will die and the battery needs to be replaced. However, since many wireless sensor nodes are deployed in harsh environments or inaccessible areas, the cost of replacing batteries is extremely high, and even impossible in some cases, which has become the main limiting factor in the development of wireless sensor networks.

[0003] To solve this problem, researchers proposed introducing solar energy harvesting (SEH) technology into wireless sensors. By collecting solar energy from the external environment and converting it into electrical energy to power wireless sensor nodes, wireless sensor networks can be self-powered and significantly extend their lifespan. In this way, sensor nodes can continue to operate without replacing batteries, avoiding node death. Wireless sensor nodes powered by solar energy harvesting technology are called solar energy harvesting wireless sensors (SEHWS). In the process of solar energy harvesting, energy prediction is a necessary prerequisite to ensure that solar energy harvesting wireless sensor nodes arrange energy distribution reasonably.

[0004] Existing energy prediction technologies mainly focus on prediction algorithms based on machine learning, deep learning and other models, as well as prediction methods based on data streams. Since EHWS is a strictly resource-constrained system, its computing power, storage space and other limitations place more stringent requirements on prediction methods. Prediction algorithms based on machine learning, deep learning and other models process and train large amounts of data. Although they have advantages in prediction effects and can achieve higher accuracy, they have large demands on computing resources and poor interpretability, and will also lead to more energy consumption. Many researchers have proposed prediction methods and algorithms based on data streams, such as EWMA, WCMA and Pro-Energy.

[0005] EWMA performs well in environments with relatively stable weather conditions, but this method is easily affected by weather changes. WCMA introduces the GAP factor based on EWMA to reduce the prediction error caused by weather changes. Pro-Energy considers a set of various energy collection modes that describe different weather conditions (clouds, sunny or rainy days) and combines the latest measurements with the weather profile closest to the current energy distribution to generate energy predictions, further improving the prediction accuracy. However, these methods are not good at capturing energy trend changes in environments with drastic or trending changes in solar energy. They are prone to lag effects and lack of sensitivity to trends when there is a trend of sustained growth or decline, resulting in low prediction accuracy. Summary of the invention

[0006] In view of this, the present invention provides a SEHWS energy prediction method and SEH wireless sensor based on double exponential moving average, which adopts the double exponential moving average prediction method to predict the energy collected by the solar energy collection wireless sensor, responds to the fluctuation of environmental energy more quickly, improves the prediction accuracy, and ensures that the wireless sensor nodes can operate stably.

[0007] The technical solution adopted by the embodiment of the present invention to solve the technical problem is:

[0008] A SEHWS energy prediction method based on double exponential moving average, comprising:

[0009] Step S1, collect and process historical energy collection data, and construct a historical D-day energy matrix E based on the current day d, where the size of E is D×N, where D represents the number of historical days, and N represents the total number of equally divided time slots in a day. In the matrix E, E(i,n) represents the energy value collected at the nth time slot on the i-th day, and E(d,n) represents the energy value collected at the nth time slot on the current day d;

[0010] Step S2, establishing a DEMA energy prediction model:

[0011]

[0012] In the formula, is the energy prediction value collected at the nth time slot of the current day d; ω is the weight factor used to control the recent energy data, ω∈(0,1); is the energy information collected by the previous time slot prediction; μ(d,n) represents the average value of the real energy information E(di,n) collected in the nth time slot of the past D days of the current day d, i∈[1,D]; B(d,n) is the correction value of the energy change trend of the nth time slot of the current day d;

[0013] Step S3, calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, where:

[0014] μ D The calculation formula for (d,n) is:

[0015]

[0016] Where n is the current time slot, i is the index variable, i∈[1,D], and E is the matrix that stores the actual collected energy;

[0017] The correction factor B(d,n) is expressed as the weighted value of the most recent energy change trend B(d,n-1) and the average energy change trend μ(d,n)-μ(d,n-1), and the expression is:

[0018] B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1))

[0019] Where γ is the weight factor used to adjust the ratio between B(d,n-1) and μ(d,n)-μ(d,n-1), γ∈(0,1), and the initial value of the iterative calculation of B(d,n) is defined as B(0,0);

[0020] Update weight factors ω and γ:

[0021] Based on the value ranges of the weight factors ω and γ, the combination of the two weight factors is brought into the DEMA energy prediction model for prediction, and the mean absolute percentage error MAPE between the true value and the predicted value is calculated to determine the weight factor combination with the lowest MAPE value;

[0022] Step S4, based on the historical energy collection data, calculate the four parameters B(d,n), μ and D(d, n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days.

[0023] Preferably, B(0,0) takes the value of 0.

[0024] Preferably, the expression of mean absolute percentage error MAPE is:

[0025]

[0026] A SEH wireless sensor comprises a microcontroller, and a data acquisition module, a solar energy collection module, a memory, a wireless communication module and a power management module connected to the microcontroller; wherein the memory stores instructions executable by the microcontroller, and the instructions are executed by the microcontroller so that the microcontroller can execute the method shown in steps S1-S4;

[0027] Data acquisition module: used to obtain environmental parameters in real time through sensors, including temperature, humidity, light, and air pressure;

[0028] Solar energy collection module: It consists of solar panels and is used to convert the collected sunlight into electrical energy;

[0029] Power management module: responsible for managing the power generated by the solar energy collection module, including:

[0030] Charging management unit: stores the electricity converted from solar energy into the battery;

[0031] Voltage regulation unit: used to supply power to each module in the SEH wireless sensor to ensure stable output voltage;

[0032] Energy monitoring unit: used to monitor the battery power and health status in real time;

[0033] Wireless communication module: used to transmit the data collected by SEH wireless sensors to the upper-level equipment through wireless signal transmission;

[0034] The memory is used to store historical energy collection data and operating programs required for the operation of the sensor;

[0035] The microcontroller is used to process historical energy collection data, and construct a historical D-day energy matrix E based on the current day d, where the size of E is D×N, N represents the total number of equally divided time slots in a day, and E(i,n) in the matrix E represents the energy value collected in the nth time slot of the i-th day, and E(d,n) represents the energy value collected in the nth time slot of the current day d;

[0036] The microcontroller is used to establish a DEMA energy prediction model:

[0037]

[0038] In the formula, is the predicted energy value collected at the nth time slot of the current day d; ω is the weight factor used to control the recent energy data, ω∈(0,1); is the energy information collected by the previous time slot prediction; μ(d,n) represents the average value of the real energy information E(di,n) collected in the nth time slot of the past D days of the current day d, i∈[1,D]; B(d,n) is the correction value of the energy change trend of the nth time slot of the current day d;

[0039] The microcontroller is used to calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, where:

[0040] μ D The calculation formula for (d,n) is:

[0041]

[0042] Where n is the current time slot, i is the index variable, i∈[1,D];

[0043] The correction factor B(d,n) is expressed as the weighted value of the most recent energy change trend B(d,n-1) and the average energy change trend μ(d,n)-μ(d,n-1), and the expression is:

[0044] B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1))

[0045] Where γ is the weight factor used to adjust the ratio between B(d,n-1) and μ(d,n)-μ(d,n-1), γ∈(0,1), and the initial value of the iterative calculation of B(d,n) is defined as B(0,0);

[0046] Update weight factors ω and γ:

[0047] Based on the value ranges of the weight factors ω and γ, the combination of the two weight factors is brought into the DEMA energy prediction model for prediction, and the mean absolute percentage error MAPE between the true value and the predicted value is calculated to determine the weight factor combination with the lowest MAPE value;

[0048] The microcontroller is used to calculate the four parameters B(d,n), μD (d, n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days.

[0049] Preferably, B(0,0) takes the value of 0.

[0050] Preferably, the expression of mean absolute percentage error MAPE is:

[0051]

[0052] It can be seen from the above technical solution that the SEHWS energy prediction method based on double exponential moving average and the SEH wireless sensor provided by the embodiment of the present invention first collect and process historical energy collection data, and build a historical D-day energy matrix E based on the current day d, where the size of E is D×N, N represents the total number of equally divided time slots in a day, and E(i,n) in the matrix E represents the energy value collected in the nth time slot of the i-th day, and E(d,n) represents the energy value collected in the nth time slot of the current day d; establish a DEMA energy prediction model; calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ; Based on the historical energy collection data, according to the four parameters B(d,n), μ D (d, n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days. The present invention adopts a double exponential moving average prediction method to predict the energy collected by the solar energy collection wireless sensor, which can respond to the fluctuation of environmental energy more quickly, improve the prediction accuracy, and ensure that the wireless sensor node can operate stably. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of the SEHWS energy prediction method based on double exponential moving average of the present invention.

[0054] Figure 2 Schematic diagram of DEMA energy prediction model. DETAILED DESCRIPTION

[0055] The technical scheme and technical effects of the present invention are further elaborated in detail below in conjunction with the accompanying drawings of the present invention.

[0056] The purpose of the present invention is to provide a solar energy collection wireless sensor energy prediction method that can reflect the energy change trend collected by the solar wireless sensor, a double exponential moving average (DEMA) prediction method, the core idea is to perform exponential smoothing on the data twice, introduce a correction factor, adjust the ratio between the correction value of the previous time slot and the average energy difference, further smooth the prediction result to better capture the trend and change in the data, respond more quickly to the fluctuation of environmental energy, reduce the prediction error, ensure that the wireless sensor node can operate stably under the condition of energy change, and make the prediction more sensitive, accurate and timely.

[0057] Please refer to Figure 1 The present invention provides a SEHWS energy prediction method based on double exponential moving average, the steps comprising:

[0058] Step S1, collect and process historical energy collection data, and construct a historical D-day energy matrix E based on the current day d, where the size of E is D×N, where D represents the number of historical energy observation days, and N represents the total number of equally divided time slots in a day. In the matrix E, E(i,n) represents the energy value collected at the nth time slot on the i-th day, and E(d,n) represents the energy value collected at the nth time slot on the current day d;

[0059] Before forecasting, each day is divided into N time slots. For example, if one time slot is chosen every half hour, that is, N = 48, in this way, the time series data is converted into a two-dimensional matrix for subsequent processing and analysis, such as Figure 2 As shown;

[0060] Step S2, establish a DEMA energy prediction model. The prediction formula of DEMA is shown in formula (1):

[0061]

[0062] In the formula, is the energy prediction value collected at the nth time slot of the current day d; ω is the weight factor used to control the recent energy data, ω∈(0,1), and they are weighted by the two weight factors ω and (1-ω). This weighted average is continuously updated with each new time slot, and its value is set between 0 and 1. Its possible values ​​are traversed, and the appropriate value is selected to improve the prediction accuracy. The most recent data point (current time slot) will have a larger weight, while the weight of the earlier data point (previous time slot) will gradually decrease; is the energy information collected by the previous time slot prediction; μ(d,n) represents the average value of the real energy information E(di,n) collected in the nth time slot of the past D days on the current day d, i∈[1,D]; B(d,n) is the correction value of the energy change trend of the nth time slot on the current day d, which is the weighted value of the most recent energy change trend and the average energy change trend;

[0063] Step S3, calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, where:

[0064] 3.1) Calculate the average value μ of the actual collected energy information of the current time slot in the past D days D (d,n), the calculation formula is:

[0065]

[0066] In the formula, d is the current date, n is the current time slot, D represents the number of days of energy observation, i is the index variable traversing the past D days, i∈[1,D], and E is the matrix that stores the actual energy collected; the energy data in time period n over the past D days will be obtained and stored in the energy matrix E, and the average value of these data will be taken. In this way, the average energy collection of the actual energy information collected in the current time period over the past D days can be obtained. As time goes by, the calculation range of the average value will continue to advance;

[0067] 3.2) The correction factor B(d,n) is expressed as the weighted value of the most recent energy change trend B(d,n-1) and the average energy change trend μ(d,n)-μ(d,n-1), and the expression is:

[0068] B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1)) (3)

[0069] Wherein, γ is the weight factor used to adjust the ratio between B(d,n-1) and μ(d,n)-μ(d,n-1), γ∈(0,1), and the weighting will also be updated with the passage of time; the initial value of the iterative calculation of B(d,n) is defined as B(0,0); B(0,0) takes the value of 0; the difference between the average energy difference in the current period and the previous period is calculated, and then the weighted average is taken with the correction value of the previous period to obtain the correction factor of the current period, which reflects the trend of the average energy change.

[0070] 3.3) Update weight factors ω and γ:

[0071] Based on the value range of weight factors ω and γ, the two weight factors are traversed, the combination of the two weight factors is brought into the DEMA energy prediction model for prediction, and the mean absolute percentage error MAPE between the true value and the predicted value is calculated to determine the weight factor combination with the lowest MAPE value; by continuously adjusting ω and γ, the optimal weight factor that achieves the best level of prediction results is selected;

[0072] The expression of mean absolute percentage error MAPE is:

[0073]

[0074] Step S4, based on the historical energy collection data (specifically, the actual energy collection information in the past few days), the four parameters B(d,n), μ and μ required for the DEMA energy prediction model are calculated according to step S3. D (d,n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days. This weighted average is continuously updated with each new time slot.

[0075] The DEMA energy prediction model is established through the above four steps. The present invention uses two exponential smoothing calculations to smooth the horizontal value and trend value of the data respectively. The smoothing operation is predicted by combining historical data and current data in a weighted moving average manner. Each new prediction value or correction value is obtained by weighted averaging the value of the previous time slot and the data of the current time slot. In the formula B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1)), the formula we use can reflect the energy change trend. Two weights ω and γ are used to perform a weighted average of the overall prediction and the correction factor. After performing two weighted moving averages, the average energy change trend can be displayed while maintaining smoothness and reducing random fluctuations in the data. This means that DEMA has a faster response speed to energy changes and can reflect the energy change trend more timely, especially when the data has an obvious upward or downward trend, making the prediction result more stable and reliable. This is particularly critical for solar energy data that may be affected by multiple factors such as weather and location in the environment.

[0076] Furthermore, the present invention provides a SEH wireless sensor, comprising a microcontroller, and a data acquisition module, a solar energy collection module, a memory, a wireless communication module and a power management module connected to the microcontroller; wherein the memory stores instructions executable by the microcontroller, and the instructions are executed by the microcontroller so that the microcontroller can execute steps S1-S4 and Figure 1The method shown;

[0077] Data acquisition module: used to obtain environmental parameters in real time through sensors. Here, environmental parameters include temperature, humidity, light, air pressure, etc.;

[0078] Solar energy collection module: mainly composed of solar panels, which are used to convert the collected sunlight into electrical energy; photovoltaic materials such as silicon crystals are usually used to improve the energy conversion efficiency;

[0079] Power management module: responsible for managing the power generated by the solar energy collection module, including:

[0080] Charging management unit: effectively stores the electricity converted from solar energy into the battery;

[0081] Voltage regulation unit: used to supply power to each module in the SEH wireless sensor to ensure stable output voltage;

[0082] Energy monitoring unit: used to monitor the battery power and health status in real time to avoid overcharging or over-discharging.

[0083] Wireless communication module: used to transmit the data collected by SEH wireless sensors to upper-level devices, such as central servers or other devices, through wireless signal transmission. Common communication technologies include Wi-Fi, LoRa, Zigbee, Bluetooth, etc. The specific choice depends on the application scenario.

[0084] A memory for storing historical energy harvesting data and operating programs required for the operation of the sensor;

[0085] The microcontroller is used to process the historical energy collection data and construct the historical D-day energy matrix E based on the current day d. The size of E is D×N, where N represents the total number of equally divided time slots in a day. In the matrix E, E(i,n) represents the energy value collected in the nth time slot on the i-th day, and E(d,n) represents the energy value collected in the nth time slot on the current day d.

[0086] Microcontroller for building DEMA energy prediction model:

[0087]

[0088] Microcontroller for calculating the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, the calculation method is similar to the above step S2;

[0089] The microcontroller is used to calculate the four parameters B(d,n), μ and D(d,n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days.

[0090] The present invention adds a smoothing calculation of the trend on the basis of the single exponential moving average, so that DEMA can better capture the trend changes of the collected solar energy data, and retain more trend information while smoothing the data, which is particularly important for solar energy collection data with certain trends. By more accurately predicting future energy changes, it can more accurately reflect the actual situation and improve the accuracy of the prediction.

[0091] The above disclosure is only the preferred embodiment of the present invention, and it is certainly not used to limit the scope of the rights of the present invention. A person skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should be included in the scope of the claims of the present invention. Therefore, the protection scope of the present invention should include any modification or equivalent alternative to the present technical solution, as long as they still fall within the purpose and scope of the present technical solution.

Claims

1. A SEHWS energy prediction method based on double exponential moving average, characterized in that: include: Step S1, collect and process historical energy collection data, and construct a historical D-day energy matrix E based on the current day d, where the size of E is D×N, N represents the total number of equally divided time slots in a day, and let E(d,n) represent the energy value collected in the nth time slot of the current day d; Step S2, establishing a DEMA energy prediction model: In the formula, is the energy prediction value collected at the nth time slot on the current day d; ω is the weight factor used to control the recent energy data, ω∈(0,1); is the energy information collected by the previous time slot prediction; μ(d,n) represents the average value of the real energy information E(di,n) collected in the nth time slot of the past D days before the current day d, i∈[1,D]; B(d,n) is the correction value of the energy change trend of the nth time slot on the current day d; Step S3, calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, where: μ D The calculation formula for (d,n) is: Where n is the current time slot, i is the index variable, i∈[1,D]; The correction factor B(d,n) is expressed as the weighted value of the most recent energy change trend B(d,n-1) and the average energy change trend μ(d,n)-μ(d,n-1), and the expression is: B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1)) Where γ is the weight factor used to adjust the ratio between B(d,n-1) and μ(d,n)-μ(d,n-1), γ∈(0,1), and the initial value of the iterative calculation of B(d,n) is defined as B(0,0); Update weight factors ω and γ: Based on the value ranges of the weight factors ω and γ, the combination of the two weight factors is brought into the DEMA energy prediction model for prediction, and the mean absolute percentage error MAPE between the true value and the predicted value is calculated to determine the weight factor combination with the lowest MAPE value; Step S4, based on the historical energy collection data, calculate the four parameters B(d,n), μ and D (d, n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days.

2. The SEHWS energy prediction method based on double exponential moving average according to claim 1, characterized in that: B(0,0) takes the value of 0.

3. The SEHWS energy prediction method based on double exponential moving average according to claim 1, characterized in that: The expression of mean absolute percentage error MAPE is:

4. A SEH wireless sensor, characterized in that: It comprises a microcontroller, and a data acquisition module, a solar energy collection module, a memory, a wireless communication module and a power management module connected to the microcontroller; wherein the memory stores instructions executable by the microcontroller, and the instructions are executed by the microcontroller so that the microcontroller can execute any of the methods described in claims 1-3; The data acquisition module is used to obtain environmental parameters in real time through sensors, and the environmental parameters include temperature, humidity, light, and air pressure; The solar energy collection module is composed of solar panels and is used to convert the collected sunlight into electrical energy; The power management module is responsible for managing the electric energy generated by the solar energy collection module, and specifically includes: Charging management unit: stores the electricity converted from solar energy into the battery; Voltage regulation unit: used to supply power to each module in the SEH wireless sensor to ensure stable output voltage; Energy monitoring unit: used to monitor the battery power and health status in real time; The wireless communication module is used to transmit the data collected by the SEH wireless sensor to the upper device through wireless signal transmission; The memory is used to store historical energy collection data and operating programs required for the operation of the sensor; The microcontroller is used to process historical energy collection data, and construct a historical D-day energy matrix E based on the current day d, where the size of E is D×N, N represents the total number of equally divided time slots in a day, and E(i,n) in the matrix E represents the energy value collected in the nth time slot of the i-th day, and E(d,n) represents the energy value collected in the nth time slot of the current day d; The microcontroller is used to establish a DEMA energy prediction model: In the formula, is the predicted energy value collected at the nth time slot of the current day d; ω is the weight factor used to control the recent energy data, ω∈(0,1); is the energy information collected by the previous time slot prediction; μ(d,n) represents the average value of the real energy information E(di,n) collected in the nth time slot of the past D days of the current day d, i∈[1,D]; B(d,n) is the correction value of the energy change trend of the nth time slot of the current day d; The microcontroller is used to calculate the parameter μ in the DEMA energy prediction model D (d,n), B(d,n), ω and γ, where: μ D The calculation formula for (d,n) is: Where n is the current time slot, i is the index variable, i∈[1,D]; The correction factor B(d,n) is expressed as the weighted value of the most recent energy change trend B(d,n-1) and the average energy change trend μ(d,n)-μ(d,n-1), and the expression is: B(d,n)=γ*B(d,n-1)+(1-γ)*(μ(d,n)-μ(d,n-1)) Where γ is the weight factor used to adjust the ratio between B(d,n-1) and μ(d,n)-μ(d,n-1), γ∈(0,1), and the initial value of the iterative calculation of B(d,n) is defined as B(0,0); Update weight factors ω and γ: Based on the value ranges of the weight factors ω and γ, the combination of the two weight factors is brought into the DEMA energy prediction model for prediction, and the mean absolute percentage error MAPE between the true value and the predicted value is calculated to determine the weight factor combination with the lowest MAPE value; The microcontroller is used to calculate the four parameters B(d,n), μ D (d, n), ω and γ are substituted into the DEMA energy prediction model, and the predicted value of the collected energy in time slot n is obtained by weighted calculation of the collected energy predicted in the previous time slot and the average value of the actual collected energy information of the current time slot n in the past few days.

5. The SEH wireless sensor according to claim 4, characterized in that: B(0,0) takes the value of 0.

6. The SEH wireless sensor according to claim 4, characterized in that: The expression of mean absolute percentage error MAPE is:

Images