Household Appliance Load Control System Based on Neural Network

Abstract

The present invention discloses a household appliance load control system based on a neural network, comprising a circuit information monitoring unit, a power performance analysis unit, a load risk assessment unit and an early warning management and control unit. Circuit information is collected by the circuit information monitoring unit, and a user power analysis model is constructed based on a neural network design. Data processing is performed by the power performance analysis unit and the load risk assessment unit to generate an early warning management signal and an appliance control signal. Then, the household appliance load control processing operation is performed by the early warning management and control unit, so as to achieve targeted management of high-risk household appliances and ensure the stability and safety of power consumption. Moreover, the user power analysis model is expanded based on the neural network, so as to achieve synchronous power monitoring and load control management of household appliances within a preset range, ensure continuous and stable operation of the power system, and achieve safe power consumption for residents and energy saving and environmental protection.

Description

Technical Field

[0001] The present invention relates to the technical field of household appliance control, and in particular to a household appliance load control system based on a neural network. Background Art

[0002] Neural network is a computational model that imitates the structure and function of neurons in the human brain. It is composed of multiple layers of neurons. Models built based on neural networks can adapt to the patterns and associations of various complex occasions. The parallel computing capability of neural networks gives them advantages in large-scale data and can significantly improve the training efficiency of the model. It is worthy of development and research in the field of power system and household appliance load monitoring.

[0003] Existing household appliances have safety and stability issues during use. When the load of an appliance is too large, the circuit is prone to overload, which may cause damage to the appliance or circuit failure. Excessive load on household appliances for a long time may cause safety hazards such as overheating and fire. In addition, excessive load on household appliances will increase energy consumption, thereby increasing household energy expenditure and causing environmental impact, which may cause the power supply system to bear higher load pressure, affecting the stability and reliability of power supply.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0005] The purpose of the present invention is to solve the safety and stability problems of existing household appliances during user use, collect circuit information through a circuit information monitoring unit, and build a user power analysis model based on a neural network design, perform data processing through a power performance analysis unit and a load risk assessment unit, generate early warning management signals and electrical appliance control signals, and then perform household appliance load control processing operations through the early warning management and control unit, thereby achieving targeted management of high-risk household appliances, ensuring the stability and safety of electricity use, and saving energy and protecting the environment.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A household appliance load control system based on a neural network includes a circuit information monitoring unit, a power performance analysis unit, a load risk assessment unit, and an early warning management and control unit, and signal connections among the circuit information monitoring unit, the power performance analysis unit, the load risk assessment unit, and the early warning management and control unit;

[0008] The circuit information monitoring unit is used to collect circuit information, wherein the circuit information includes the power consumption parameters of the user bus and the power consumption parameters of the household appliances. The power consumption bus is preset to electrically connect N0 household appliances, and the information collection period Tc is set to collect the circuit information regularly;

[0009] The power consumption performance analysis unit is used to analyze circuit information: first, by analyzing the power consumption of N0 household appliances and the power supply of the user bus, the power conversion efficiency is compared and measured, so as to evaluate the household power consumption status, and then analyze the safety and stability of power consumption, so as to comprehensively evaluate the power consumption performance;

[0010] The load risk assessment unit is used to deeply analyze power consumption performance: by extending the power consumption performance, setting the power consumption performance risk interval, determining the risk level of household appliance load, and providing feedback analysis to locate high-risk appliances, generating corresponding early warning management signals and appliance control signals;

[0011] The early warning management and control unit is used to receive signals and perform corresponding household appliance load control processing operations.

[0012] Furthermore, a user power analysis model is constructed based on a neural network design, and data processing is performed through a power performance analysis unit and a load risk assessment unit. The construction process of the user power analysis model is as follows:

[0013] Sa: Construct a circuit information analysis model to perform preliminary analysis and processing on the circuit information:

[0014] Sa-1: Construct an electric parameter analysis model to preprocess circuit information;

[0015] Sa-2: Analyze and obtain the power conversion efficiency ηz and circuit energy consumption index Zh;

[0016] Sa-3: Generate household electricity consumption evaluation coefficient Xjd by combining the power conversion efficiency ηz with the circuit energy consumption index Zh;

[0017] Sb: Construct an electricity consumption curve analysis model to analyze the electricity consumption performance of the household electricity consumption evaluation coefficient Xjd:

[0018] Sb-1: construct a dynamic curve diagram S0 of household electricity evaluation coefficient Xjd-information collection period Tc;

[0019] Sb-2: Construct a fluctuation analysis model to perform stability analysis on the curve S0 and generate the curve stability coefficient φs;

[0020] Sb-3: Construct a risk analysis model to conduct safety analysis on curve S0 and generate the curve safety factor δs;

[0021] Sb-4: Generate the power performance evaluation index Zxn by combining the curve stability coefficient φs and the curve safety coefficient δs;

[0022] Sc: Construct a performance in-depth analysis model to conduct in-depth analysis of the power consumption performance evaluation index Zxn:

[0023] Sc-1: Set the power performance risk interval and generate corresponding early warning management signals through interval comparison;

[0024] Sc-2: Set the threshold of the early warning management signal level and determine whether to conduct feedback analysis to locate high-risk home appliances through threshold comparison;

[0025] Sc-3: Set a threshold value of the power consumption coefficient Xqx of the household appliance Q, and determine whether the household appliance Q is a high-risk appliance by comparing the threshold value. If the household appliance Q is determined to be a high-risk appliance, generate and output a corresponding appliance control signal.

[0026] Furthermore, the specific process of collecting circuit information is as follows:

[0027] The power consumption parameters include voltage, current and power consumption. Any user bus is marked as H. The voltage of user bus H is marked as Uh, the current is marked as Ih, and the power consumption is marked as Wh.

[0028] It is assumed that the user bus H is electrically connected to N0 household appliances, and any household appliance is marked as Q. The voltage, current and power consumption of the household appliance Q are Uq, Iq and Wq, respectively.

[0029] Furthermore, the specific process of constructing the circuit information analysis model is as follows:

[0030] A1: First, build an electricity parameter analysis model to analyze the circuit information:

[0031] A1-1: The input information of the power consumption parameter analysis model is the voltage U0, current I0 and power consumption W0 within the information collection period Tc;

[0032] A1-2: By combining the voltage U0 with the current I0, the circuit power P0 in the current information collection period Tc is obtained;

[0033] A1-3: Then, by combining the circuit power P0 and the power consumption W0, corresponding weight factor coefficients are assigned to the circuit power P0 and the power consumption W0, so as to obtain and output the circuit energy index Zn;

[0034] A2: Substitute the circuit information into the power parameter analysis model for preprocessing:

[0035] A2-1: First input the power consumption parameters of the user bus H into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the circuit energy supply index Zg of the user bus H;

[0036] A2-2: Input the power consumption parameters of the household appliance Q into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the power consumption coefficient Xqx of the household appliance Q;

[0037] A3: Then, by combining the power consumption coefficients Xqx of N0 household appliances, the circuit energy consumption index Zh of the household appliances is obtained;

[0038] A4: Then, the power conversion efficiency ηz is obtained by combining the circuit energy consumption index Zh of the household appliance and the circuit energy supply index Zg of the user bus H;

[0039] A5: By combining the electric energy conversion efficiency ηz with the circuit energy consumption index Zh, the household electricity consumption evaluation coefficient Xjd is generated to evaluate the household electricity consumption status.

[0040] Furthermore, the specific process of constructing the electricity consumption curve analysis model is as follows:

[0041] A6-1: With the household electricity consumption evaluation coefficient Xjd as the ordinate and the information collection period Tc as the abscissa, a dynamic curve graph S0 of the household electricity consumption evaluation coefficient Xjd-information collection period Tc is constructed;

[0042] A6-1-1: Collect all points and coordinates on curve S0. The default curve S0 has n0 points. Mark any point on curve S0 as i. The coordinates of point i are (Xi, Yi). Mark the point closest to point i as j (Xj, Yj).

[0043] A6-1-2: Then calculate the slope Ki of the acquisition point i;

[0044] A6-2: Construct a fluctuation analysis model: preset the input set M, process the data through the fluctuation analysis model, and output the element mean JZ and the element fluctuation coefficient σm;

[0045] A6-3: Integrate the ordinates of n0 points into set M1, integrate the slopes of n0 points into set M2, substitute sets M1 and M2 into the fluctuation analysis model in turn, and output the corresponding element means and element fluctuation coefficients;

[0046] Among them, the element mean and element fluctuation coefficient outputted by the set M1 after being processed by the fluctuation analysis model are marked as the overall value of the curve Yz and the curve fluctuation coefficient σz respectively; the element mean and element fluctuation coefficient outputted by the set M2 after being processed by the fluctuation analysis model are marked as the curve increase rate Ks and the increase rate fluctuation coefficient σk respectively;

[0047] A6-4: Perform stability analysis on the curve: generate the curve stability coefficient φs by combining the rate fluctuation coefficient σk and the curve fluctuation coefficient σz;

[0048] A6-5: Constructing a risk analysis model: Inputting curve S0 into the risk analysis model, setting the threshold of the household electricity consumption evaluation coefficient Xjd, determining whether the household electricity consumption status is qualified, and extracting all the curve segments of the household status that are unqualified in curve S0, and then performing safety analysis on the curve to generate the curve safety coefficient δs;

[0049] A6-6: The power performance evaluation index Zxn is generated by combining the curve stability coefficient φs and the curve safety coefficient δs.

[0050] Furthermore, the specific process of constructing the fluctuation analysis model is as follows:

[0051] A6-2-1: Mark the input information as a set M, which contains n0 element values, and mark any element value as m;

[0052] A6-2-2: Obtain the element mean JZ by taking the average value of n0 element values;

[0053] A6-2-3: Then find the standard deviation of n0 element values ​​and obtain the element fluctuation coefficient σm;

[0054] A6-2-4: Output element mean JZ and element fluctuation coefficient σm.

[0055] Furthermore, the specific process of building a risk analysis model is as follows:

[0056] A6-5-1: Obtain the total electricity consumption time Ta by measuring the horizontal coordinate difference between the initial point and the end point of curve S0;

[0057] A6-5-2: Set the threshold of the household electricity consumption evaluation coefficient Xjd to JD. When the household electricity consumption evaluation coefficient Xjd is lower than the threshold JD, the household electricity consumption status is judged to be unqualified; otherwise, the household electricity consumption status is judged to be qualified;

[0058] A6-5-3: Extract all curve segments below the threshold JD in the curve S0. There are n1 curve segments in advance, and mark any curve segment as L.

[0059] A6-5-301: Obtain the duration Tl of the curve by measuring the difference in the horizontal coordinates between the initial point and the end point of the curve segment L;

[0060] A6-5-302: Obtain the duration of unqualified electricity consumption Tb by measuring the sum of the curve durations of n1 curve segments;

[0061] A6-5-4: Then, the proportion of safe electricity use time Te is obtained by combining the unqualified electricity use time Tb with the total electricity use time Ta;

[0062] A6-5-5: Conduct a safety analysis on the curve: generate the curve safety factor δs by combining the safe electricity use time ratio Te with the overall value of the curve Yz.

[0063] Furthermore, a performance in-depth analysis model is constructed to conduct in-depth analysis of power consumption performance. The specific process is as follows:

[0064] By setting the power consumption performance risk interval for the power consumption performance evaluation index Zxn, there are n2 power consumption performance risk intervals preset, and any power consumption performance risk interval is marked as Rf;

[0065] Then, perform interval comparison: when the power consumption performance evaluation index Zxn is within the power consumption performance risk interval Rf, the risk level of household appliance load is determined to be level f, and a level f early warning management signal is generated and output;

[0066] A threshold value FX is set for the household appliance load risk level f. When the household appliance load risk level f is higher than the threshold value FX, feedback analysis is performed to locate high-risk appliances: by setting the threshold value of the energy consumption coefficient Xqx of the household appliance Q to XHq, when the energy consumption coefficient Xqx exceeds the threshold value XHq, the household appliance Q is determined to be a high-risk appliance, and a corresponding appliance control signal is generated and output.

[0067] Furthermore, the specific operation process of the early warning management and control unit is as follows:

[0068] When receiving an early warning management signal, immediately edit the corresponding early warning prompt text and present it visually, so as to prompt the back-end management personnel to take corresponding measures through the early warning prompt text;

[0069] When an appliance control signal is received, dynamic control operations are performed on the corresponding high-risk home appliances, and a corresponding dynamic control plan is set for the home appliance Q based on the user's behavior habits.

[0070] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0071] The present invention collects circuit information through a circuit information monitoring unit, and constructs a user power analysis model based on a neural network design, performs data processing through a power performance analysis unit and a load risk assessment unit, generates an early warning management signal and an electrical appliance control signal, and then performs household appliance load control processing operations through the early warning management control unit, thereby achieving targeted management of high-risk household appliances and ensuring the stability and safety of electricity use;

[0072] The present invention constructs a circuit information analysis model to perform preliminary analysis and processing on the circuit information, thereby evaluating the household electricity consumption status, and then constructs a power consumption curve analysis model to analyze the safety and stability of electricity consumption, thereby comprehensively evaluating the power consumption performance, realizing timely early warning management of household electricity consumption, and then constructing a performance depth analysis model to achieve accurate positioning of high-risk household appliances, thereby performing targeted dynamic regulation to avoid peak loads and ensure that the user's household circuit load is within the loadable range, thereby ensuring the stability and safety of electricity consumption;

[0073] The present invention expands the user power analysis model based on a neural network, and can realize synchronous power monitoring and load regulation management of household appliances within a preset range, ensuring continuous and stable operation of the power system and achieving safe, energy-saving and environmentally friendly electricity use for residents. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 The overall module schematic diagram of the present invention is shown;

[0075] Figure 2 The overall process diagram of the present invention is shown;

[0076] Figure 3 A module schematic diagram of the user power analysis model of the present invention is shown. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0078] Embodiment 1:

[0079] like Figure 1-3 As shown, the household appliance load control system based on neural network includes a circuit information monitoring unit, a power performance analysis unit, a load risk assessment unit and an early warning management and control unit, and the circuit information monitoring unit, the power performance analysis unit, the load risk assessment unit and the early warning management and control unit are connected by signals;

[0080] A neural network is a computational model that mimics the structure and function of neurons in the human brain. It is composed of multiple layers of neurons. When using neural networks for model design, it can be expanded by adjusting the network structure, activation function, loss function, etc. to meet the needs of different problems, such as adding more layers, the number of neurons, and the way of connecting layers, etc., to improve the performance and generalization ability of the model.

[0081] Among them, the user power analysis model is constructed based on the neural network design, and data processing is performed through the power performance analysis unit and the load risk assessment unit. The construction process of the user power analysis model is as follows:

[0082] Sa: Construct a circuit information analysis model to analyze and process the circuit information:

[0083] Sa-1: Construct an electric parameter analysis model to preprocess circuit information;

[0084] Sa-2: Analyze and obtain the power conversion efficiency ηz and circuit energy consumption index Zh;

[0085] Sa-3: Generate household electricity consumption evaluation coefficient Xjd by combining the power conversion efficiency ηz with the circuit energy consumption index Zh;

[0086] Sb: Construct an electricity consumption curve analysis model to analyze the electricity consumption performance of the household electricity consumption evaluation coefficient Xjd:

[0087] Sb-1: construct a dynamic curve diagram S0 of household electricity evaluation coefficient Xjd-information collection period Tc;

[0088] Sb-2: Construct a fluctuation analysis model to perform stability analysis on the curve S0 and generate the curve stability coefficient φs;

[0089] Sb-3: Construct a risk analysis model to conduct safety analysis on curve S0 and generate the curve safety factor δs;

[0090] Sb-4: Generate the power performance evaluation index Zxn by combining the curve stability coefficient φs and the curve safety coefficient δs;

[0091] Sc: Construct a performance in-depth analysis model to conduct in-depth analysis of the power consumption performance evaluation index Zxn:

[0092] Sc-1: Set the power performance risk interval and generate corresponding early warning management signals through interval comparison;

[0093] Sc-2: Set the threshold of the early warning management signal level and determine whether to conduct feedback analysis to locate high-risk home appliances through threshold comparison;

[0094] Sc-3: Set the threshold of the power consumption coefficient Xqx of the household appliance Q, and determine whether the household appliance Q is a high-risk appliance by comparing the threshold. If the household appliance Q is determined to be a high-risk appliance, generate and output the corresponding appliance control signal;

[0095] The specific working steps of this system are as follows:

[0096] S1: The circuit information monitoring unit collects circuit information, wherein the circuit information includes power consumption parameters of the user bus and power consumption parameters of household appliances. The power consumption bus is preset to electrically connect N0 household appliances, and an information collection period Tc is set to collect circuit information regularly;

[0097] The specific process of collecting circuit information is as follows:

[0098] The power consumption parameters include voltage, current and power consumption. Any user bus is marked as H. The voltage of user bus H is marked as Uh, the current is marked as Ih, and the power consumption is marked as Wh.

[0099] It is assumed that the user bus H is electrically connected to N0 household appliances, and any household appliance is marked as Q. The voltage, current and power consumption of household appliance Q are Uq, Iq and Wq, respectively.

[0100] S2: The power performance analysis unit is used to analyze circuit information: first, by analyzing the power consumption of N0 household appliances and the power supply of the user bus, the power conversion efficiency is compared and measured, so as to evaluate the household power consumption status, and then analyze the safety and stability of power consumption, so as to comprehensively evaluate the power performance;

[0101] Among them, the specific process of building a circuit information analysis model is:

[0102] A1: First, build an electricity parameter analysis model to analyze the circuit information. The specific process is as follows:

[0103] A1-1: The input information of the power consumption parameter analysis model is the voltage U0, current I0 and power consumption W0 within the information collection period Tc;

[0104] A1-2: By combining the voltage U0 with the current I0, the circuit power P0 in the current information collection period Tc is obtained: P0 = I0*I0;

[0105] A1-3: Then, by combining the circuit power P0 and the power consumption W0, corresponding weight factor coefficients are assigned to the circuit power P0 and the power consumption W0, so as to obtain and output the circuit energy index Zn:

[0106] Zn=ρ1*P0+ρ2*W0, where ρ1 and ρ2 are weight factor coefficients of circuit power P0 and power consumption W0, respectively, and both ρ1 and ρ2 are greater than 0; when the circuit power P0 and power consumption W0 are higher, the circuit energy index Zn is higher;

[0107] A2: Substitute the circuit information into the power parameter analysis model for preprocessing:

[0108] A2-1: First input the power consumption parameters of the user bus H into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the circuit energy supply index Zg of the user bus H;

[0109] A2-2: Input the power consumption parameters of the household appliance Q into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the power consumption coefficient Xqx of the household appliance Q;

[0110] The specific process of preprocessing is as follows:

[0111] By combining the voltage Uh with the current Ih, the circuit power Ph of the user bus H in the current information collection period Tc is obtained, and then by combining the circuit power Ph with the power consumption Wh, the circuit energy supply index Zg of the user bus H is obtained: Zg = α1*Ph+α2*Wh;

[0112] Among them, α1 and α2 are weight factor coefficients of circuit power Ph and power consumption Wh respectively, and α1 and α2 are both greater than 0. The weight factor coefficients are obtained by calculating a large amount of experimental data and then preset. When the circuit power Ph and power consumption Wh are higher, the circuit energy supply index Zg of user bus H is higher, indicating that the circuit energy supply of user bus H is more.

[0113] By combining the voltage Uq with the current Iq, the circuit power Pq of the household appliance Q in the current information collection period Tc is obtained, and then by combining the circuit power Pq with the power consumption Wq, the power consumption coefficient Xqx of the household appliance Q is obtained: Xqx = β1*Pq + β2*Wq;

[0114] Among them, β1 and β2 are weight factor coefficients of circuit power Pq and power consumption Wq, respectively, and both β1 and β2 are greater than 0; when the circuit power Pq and power consumption Wq are higher, the power consumption coefficient Xqx is higher, indicating that the household appliance Q consumes more power;

[0115] A3: Then, by combining the power consumption coefficients Xqx of N0 household appliances, the circuit energy consumption index Zh of the household appliances is obtained:

[0116] Among them, εq is the conversion coefficient of the power consumption coefficient Xqx, and εq is greater than 0; when the power consumption coefficient Xqx is higher, the circuit energy consumption index Zh of the household appliance is higher, which means that the more power the household appliance Q consumes, the more comprehensive power consumption of N0 household appliances;

[0117] A4: Then, the energy conversion efficiency ηz is obtained by combining the circuit energy consumption index Zh of the household appliance and the circuit energy supply index Zg of the user bus H:

[0118] The higher the power conversion efficiency ηz is, the lower the loss in the power transmission process is, and the higher the effective utilization rate of power is;

[0119] A5: By combining the power conversion efficiency ηz with the circuit energy consumption index Zh, the household electricity consumption evaluation coefficient Xjd is generated:

[0120] Among them, ω1 and ω2 are weight factor coefficients of the power conversion efficiency ηz and the circuit energy consumption index Zh, respectively, and ω1 and ω2 are both greater than 0; when the power conversion efficiency ηz is higher and the circuit energy consumption index Zh is lower, the household electricity consumption evaluation coefficient Xjd is higher, indicating that the household electricity consumption status is better;

[0121] A6: Construct an electricity consumption curve analysis model to analyze the electricity consumption performance of the household electricity consumption evaluation coefficient Xjd. The specific process is as follows:

[0122] A6-1: With the household electricity consumption evaluation coefficient Xjd as the ordinate and the information collection period Tc as the abscissa, a dynamic curve graph S0 of the household electricity consumption evaluation coefficient Xjd-information collection period Tc is constructed;

[0123] A6-1-1: Collect all points and coordinates on curve S0. The default curve S0 has n0 points. Mark any point on curve S0 as i. The coordinates of point i are (Xi, Yi). Mark the point closest to point i as j (Xj, Yj).

[0124] A6-1-2: Then calculate the slope Ki of point i:

[0125] A6-2: Construct a fluctuation analysis model and process the data of the ordinate and slope of the curve. The specific process is as follows:

[0126] A6-2-1: Mark the input information as a set M, which contains n0 element values, and mark any element value as m;

[0127] A6-2-2: Obtain the element mean JZ by taking the average value of n0 element values;

[0128] A6-2-3: Then find the standard deviation of n0 element values ​​and obtain the element fluctuation coefficient σm;

[0129] A6-2-4: Output element mean JZ and element fluctuation coefficient σm;

[0130] A6-3: Integrate the ordinates of n0 points into set M1, integrate the slopes of n0 points into set M2, substitute them into the fluctuation analysis model in sequence, and output the corresponding element mean and element fluctuation coefficient;

[0131] Among them, the element mean and element fluctuation coefficient outputted by the set M1 after being processed by the fluctuation analysis model are marked as the overall value of the curve Yz and the curve fluctuation coefficient σz respectively; the element mean and element fluctuation coefficient outputted by the set M2 after being processed by the fluctuation analysis model are marked as the curve increase rate Ks and the increase rate fluctuation coefficient σk respectively;

[0132] A6-3-1: Obtain the overall value Yz of the curve by taking the average value of the ordinates of n0 points: Then calculate the standard deviation and obtain the curve fluctuation coefficient σz:

[0133] A6-3-2: Obtain the curve increase rate Ks by taking the average slope of n0 points: Then calculate the standard deviation and obtain the rate of increase fluctuation coefficient σk:

[0134] A6-4: Perform stability analysis on the curve: By combining the rate fluctuation coefficient σk and the curve fluctuation coefficient σz, the curve stability coefficient φs is generated:

[0135] Among them, υ1 and υ2 are the weight coefficients of the rate fluctuation coefficient σk and the curve fluctuation coefficient σz, respectively, and υ1 and υ2 are both greater than 0; when the rate fluctuation coefficient σk and the curve fluctuation coefficient σz are lower, the curve stability coefficient φs is higher, indicating that the power consumption stability is better;

[0136] A6-5: Construct a risk analysis model. The specific process is as follows:

[0137] A6-5-1: Obtain the total electricity consumption time Ta by measuring the horizontal coordinate difference between the initial point and the end point of curve S0;

[0138] A6-5-2: Set the threshold of the household electricity consumption evaluation coefficient Xjd to JD. When the household electricity consumption evaluation coefficient Xjd is lower than the threshold JD, the household electricity consumption status is judged to be unqualified; otherwise, the household electricity consumption status is judged to be qualified;

[0139] A6-5-3: Extract all curve segments below the threshold JD in the curve S0. There are n1 curve segments in advance, and mark any curve segment as L.

[0140] A6-5-301: Obtain the duration Tl of the curve by measuring the difference in the horizontal coordinates between the initial point and the end point of the curve segment L;

[0141] A6-5-302: Obtain the duration of unqualified electricity consumption Tb by measuring the sum of the curve durations of n1 curve segments;

[0142] A6-5-4: Then, by combining the unqualified electricity consumption time Tb with the total electricity consumption time Ta, we can obtain the proportion of safe electricity consumption time Te:

[0143] A6-5-5: Conduct a safety analysis on the curve: By combining the safe electricity use time ratio Te with the overall value of the curve Yz, generate the curve safety factor δs:

[0144] Among them, γ1 and γ2 are the weight factor coefficients of the safe electricity use time ratio Te and the overall value of the curve Yz, and both γ1 and γ2 are greater than 0; when the safe electricity use time ratio Te is higher and the overall value of the curve Yz is higher, the curve safety coefficient δs is higher, indicating that the electricity safety is better;

[0145] A6-6: By combining the curve stability coefficient φs and the curve safety coefficient δs, the power performance evaluation index Zxn is generated: Zxn=u1*φs+u2*δs;

[0146] Among them, μ1 and μ2 are weight factor coefficients of the curve stability coefficient φs and the curve safety coefficient δs, respectively, and both μ1 and μ2 are greater than 0; when the curve stability coefficient φs and the curve safety coefficient δs are higher, the power performance evaluation index Zxn is higher, indicating that the power performance is better;

[0147] S3: Load risk assessment unit deeply analyzes power consumption performance: by extending the power consumption performance, setting the power consumption performance risk interval, determining the risk level of household appliance load, and providing feedback analysis to locate high-risk appliances, generating corresponding early warning management signals and appliance control signals;

[0148] Among them, the specific process of in-depth analysis of power performance is:

[0149] By setting the power consumption performance risk interval for the power consumption performance evaluation index Zxn, there are n2 power consumption performance risk intervals preset, and marking any power consumption performance risk interval as Rf, where f refers to the sequence number of the risk interval, 0<f≤n2 and f is a natural number;

[0150] Then, perform interval comparison: when the power consumption performance evaluation index Zxn is within the power consumption performance risk interval Rf, the risk level of household appliance load is determined to be level f, and a level f early warning management signal is generated and output;

[0151] Set the threshold FX of the household appliance load risk level f. When the household appliance load risk level f is higher than the threshold FX, feedback analysis is performed to locate high-risk appliances:

[0152] The threshold value of the power consumption coefficient Xqx of the household appliance Q is set to XHq. When the power consumption coefficient Xqx exceeds the threshold value XHq, the household appliance Q is determined to be a high-risk household appliance, and a corresponding appliance control signal is generated and output;

[0153] S4: The early warning management and control unit receives the signal and performs corresponding processing operations;

[0154] When receiving an early warning management signal, immediately edit the corresponding early warning prompt text and present it visually, so as to prompt the back-end management personnel to take corresponding measures through the early warning prompt text;

[0155] When receiving an appliance control signal, a dynamic control operation is performed on the corresponding high-risk home appliance, and a corresponding dynamic control scheme is set for the home appliance Q according to the user's behavior habits. The dynamic control scheme includes a low power consumption mode and a power-off mode. For example, if the user sets a refrigerator as a home appliance that cannot be powered off, the low power consumption mode is applied to the home appliance that cannot be powered off; if the user sets a lamp as a home appliance that allows power off, the power-off mode is applied to the home appliance that allows power off;

[0156] By accurately locating high-risk home appliances and regulating them, it is possible to avoid turning on high-power appliances at the same time, thereby avoiding peak loads and ensuring that the user's home circuit load is within the loadable range, thereby ensuring the stability and safety of electricity use;

[0157] It should be noted that based on the neural network, the user power analysis model can be extended to the N1 user households in a building to conduct building power analysis and household appliance load regulation, thereby realizing building power consumption monitoring and management, and then extended to the N2 buildings in a community to conduct community power analysis and household appliance load regulation, thereby realizing community power consumption monitoring and management, and so on, ultimately realizing the synchronous power monitoring and load regulation management of household appliances within a preset range, ensuring the continuous and stable operation of the power system and realizing the safety of residents' power consumption.

[0158] In summary, the present invention collects circuit information through the circuit information monitoring unit, and constructs a user power analysis model based on a neural network design, performs data processing through the power performance analysis unit and the load risk assessment unit, generates an early warning management signal and an electrical appliance control signal, and then performs household appliance load control processing operations through the early warning management control unit, thereby achieving targeted management of high-risk household appliances and ensuring the stability and safety of electricity use;

[0159] The present invention constructs a circuit information analysis model to perform preliminary analysis and processing on the circuit information, thereby evaluating the household electricity consumption status, and then constructs a power consumption curve analysis model to analyze the safety and stability of electricity consumption, thereby comprehensively evaluating the power consumption performance, realizing timely early warning management of household electricity consumption, and then constructing a performance depth analysis model to achieve accurate positioning of high-risk household appliances, thereby performing targeted dynamic regulation to avoid peak loads and ensure that the user's household circuit load is within the loadable range, thereby ensuring the stability and safety of electricity consumption;

[0160] The present invention expands the user power analysis model based on a neural network, and can realize synchronous power monitoring and load regulation management of household appliances within a preset range, ensuring continuous and stable operation of the power system and achieving safe electricity use for residents.

[0161] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technical personnel in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0162] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0163] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A household appliance load control system based on a neural network, characterized by: It includes a circuit information monitoring unit, a power performance analysis unit, a load risk assessment unit and an early warning management and control unit, and signal connections between the circuit information monitoring unit, the power performance analysis unit, the load risk assessment unit and the early warning management and control unit; The circuit information monitoring unit is used to collect circuit information, wherein the circuit information includes the power consumption parameters of the user bus and the power consumption parameters of the household appliances. It is preset that the power consumption bus is electrically connected with N0 household appliances, and the information collection period Tc is set to collect the circuit information regularly; The power performance analysis unit is used to analyze circuit information: first, by analyzing the power consumption of N0 household appliances and the power supply of the user bus, the power conversion efficiency is compared and measured, so as to evaluate the household power consumption status, and then analyze the safety and stability of power consumption, so as to comprehensively evaluate the power performance; The load risk assessment unit is used to deeply analyze power consumption performance: by extending the power consumption performance, setting the power consumption performance risk interval, determining the risk level of household appliance load, and providing feedback analysis to locate high-risk appliances, generating corresponding early warning management signals and appliance control signals; The early warning management and control unit is used to receive signals and perform corresponding household appliance load control processing operations; The user power analysis model is constructed based on the neural network design, and data processing is performed through the power performance analysis unit and the load risk assessment unit. The construction process of the user power analysis model is as follows: Sa: Construct a circuit information analysis model to perform preliminary analysis and processing on the circuit information: Sa-1: Construct an electric parameter analysis model to preprocess circuit information; Sa-2: Analyze and obtain the power conversion efficiency ηz and circuit energy consumption index Zh; Sa-3: Generate household electricity consumption evaluation coefficient Xjd by combining the power conversion efficiency ηz with the circuit energy consumption index Zh; Sb: Construct an electricity consumption curve analysis model to analyze the electricity consumption performance of the household electricity consumption evaluation coefficient Xjd: Sb-1: construct a dynamic curve diagram S0 of the household electricity consumption evaluation coefficient Xjd information collection period Tc; Sb-2: Construct a fluctuation analysis model to perform stability analysis on the curve S0 and generate the curve stability coefficient φs; Sb-3: Construct a risk analysis model to conduct safety analysis on curve S0 and generate the curve safety factor δs; Sb-4: Generate the power performance evaluation index Zxn by combining the curve stability coefficient φs and the curve safety coefficient δs; Sc: Construct a performance in-depth analysis model to conduct in-depth analysis of the power consumption performance evaluation index Zxn: Sc-1: Set the power performance risk interval and generate corresponding early warning management signals through interval comparison; Sc-2: Set the threshold of the early warning management signal level and determine whether to conduct feedback analysis to locate high-risk home appliances through threshold comparison; Sc-3: Set a threshold value of the power consumption coefficient Xqx of the household appliance Q, and determine whether the household appliance Q is a high-risk appliance by comparing the threshold value. If the household appliance Q is determined to be a high-risk appliance, generate and output a corresponding appliance control signal.

2. The household appliance load control system based on neural network according to claim 1 is characterized in that: The specific process of collecting circuit information is as follows: The power consumption parameters include voltage, current and power consumption. Any user bus is marked as H. The voltage of user bus H is marked as Uh, the current is marked as Ih, and the power consumption is marked as Wh. It is assumed that the user bus H is electrically connected to N0 household appliances, and any household appliance is marked as Q. The voltage, current and power consumption of the household appliance Q are Uq, Iq and Wq, respectively.

3. The household appliance load control system based on neural network according to claim 2 is characterized in that: The specific process of building a circuit information analysis model is as follows: A1: First, build an electricity parameter analysis model to analyze the circuit information: A1-1: The input information of the power consumption parameter analysis model is the voltage U0, current I0 and power consumption W0 within the information collection period Tc; A1-2: By combining the voltage U0 with the current I0, the circuit power P0 in the current information collection period Tc is obtained; A1-3: Then, by combining the circuit power P0 and the power consumption W0, corresponding weight factor coefficients are assigned to the circuit power P0 and the power consumption W0, so as to obtain and output the circuit energy index Zn; A2: Substitute the circuit information into the power parameter analysis model for preprocessing: A2-1: First, input the power consumption parameters of the user bus H into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the circuit energy supply index Zg of the user bus H; A2-2: Input the power consumption parameters of the household appliance Q into the power consumption parameter analysis model, obtain the corresponding circuit energy index Zn and mark it as the power consumption coefficient Xqx of the household appliance Q; A3: Then, by combining the power consumption coefficients Xqx of N0 household appliances, the circuit energy consumption index Zh of the household appliances is obtained; A4: Then, the energy conversion efficiency ηz is obtained by combining the circuit energy consumption index Zh of the household appliance and the circuit energy supply index Zg of the user bus H; A5: By combining the electric energy conversion efficiency ηz with the circuit energy consumption index Zh, the household electricity consumption evaluation coefficient Xjd is generated to evaluate the household electricity consumption status.

4. The household appliance load control system based on neural network according to claim 3 is characterized in that: The specific process of constructing the electricity consumption curve analysis model is as follows: A6-1: With the household electricity consumption evaluation coefficient Xjd as the ordinate and the information collection period Tc as the abscissa, a dynamic curve graph S0 of the household electricity consumption evaluation coefficient Xjd-information collection period Tc is constructed; A6-1-1: Collect all points and coordinates on curve S0. The default curve S0 has n0 points. Mark any point on curve S0 as i. The coordinates of point i are (Xi, Yi). Mark the point closest to point i as j (Xj, Yj). A6-1-2: Then calculate the slope Ki of the acquisition point i; A6-2: Construct a fluctuation analysis model: preset the input set M, process the data through the fluctuation analysis model, and output the element mean JZ and the element fluctuation coefficient σm; A6-3: Integrate the ordinates of n0 points into set M1, integrate the slopes of n0 points into set M2, substitute sets M1 and M2 into the fluctuation analysis model in turn, and output the corresponding element means and element fluctuation coefficients; Among them, the element mean and element fluctuation coefficient outputted by the set M1 after being processed by the fluctuation analysis model are marked as the overall value of the curve Yz and the curve fluctuation coefficient σz respectively; the element mean and element fluctuation coefficient outputted by the set M2 after being processed by the fluctuation analysis model are marked as the curve increase rate Ks and the increase rate fluctuation coefficient σk respectively; A6-4: Perform stability analysis on the curve: generate the curve stability coefficient φs by combining the rate fluctuation coefficient σk and the curve fluctuation coefficient σz; A6-5: Constructing a risk analysis model: Inputting curve S0 into the risk analysis model, setting the threshold of the household electricity consumption evaluation coefficient Xjd, determining whether the household electricity consumption status is qualified, and extracting all the curve segments of the household status that are unqualified in curve S0, and then performing safety analysis on the curve to generate the curve safety coefficient δs; A6-6: The power performance evaluation index Zxn is generated by combining the curve stability coefficient φs and the curve safety coefficient δs.

5. The household appliance load control system based on neural network according to claim 4 is characterized in that: The specific process of constructing a volatility analysis model is as follows: A6-2-1: Mark the input information as a set M, which contains n0 element values, and mark any element value as m; A6-2-2: Obtain the element mean JZ by taking the average value of n0 element values; A6-2-3: Then find the standard deviation of n0 element values ​​and obtain the element fluctuation coefficient σm; A6-2-4: Output element mean JZ and element fluctuation coefficient σm.

6. The household appliance load control system based on neural network according to claim 5, characterized in that: The specific process of building a risk analysis model is as follows: A6-5-1: Obtain the total electricity consumption time Ta by measuring the horizontal coordinate difference between the initial point and the end point of curve S0; A6-5-2: Set the threshold of the household electricity consumption evaluation coefficient Xjd to JD. When the household electricity consumption evaluation coefficient Xjd is lower than the threshold JD, the household electricity consumption status is judged to be unqualified; otherwise, the household electricity consumption status is judged to be qualified; A6-5-3: Extract all curve segments below the threshold JD in the curve S0. There are n1 curve segments in advance, and mark any curve segment as L. A6-5-301: Obtain the duration Tl of the curve by measuring the difference in the horizontal coordinates between the initial point and the end point of the curve segment L; A6-5-302: Obtain the duration of unqualified electricity consumption Tb by measuring the sum of the curve durations of n1 curve segments; A6-5-4: Then, the proportion of safe electricity use time Te is obtained by combining the unqualified electricity use time Tb with the total electricity use time Ta; A6-5-5: Conduct a safety analysis on the curve: generate the curve safety factor δs by combining the safe electricity use time ratio Te with the overall value of the curve Yz.

7. The household appliance load control system based on neural network according to claim 6, characterized in that: Construct a performance in-depth analysis model to conduct in-depth analysis of power consumption performance. The specific process is as follows: By setting the power consumption performance risk interval for the power consumption performance evaluation index Zxn, there are n2 power consumption performance risk intervals preset, and any power consumption performance risk interval is marked as Rf; Then, perform interval comparison: when the power consumption performance evaluation index Zxn is within the power consumption performance risk interval Rf, the risk level of household appliance load is determined to be level f, and a level f early warning management signal is generated and output; A threshold value FX is set for the household appliance load risk level f. When the household appliance load risk level f is higher than the threshold value FX, feedback analysis is performed to locate high-risk appliances: by setting the threshold value of the energy consumption coefficient Xqx of the household appliance Q to XHq, when the energy consumption coefficient Xqx exceeds the threshold value XHq, the household appliance Q is determined to be a high-risk appliance, and a corresponding appliance control signal is generated and output.

8. The household appliance load control system based on neural network according to claim 7, characterized in that: The specific operation process of the early warning management and control unit is as follows: When receiving an early warning management signal, immediately edit the corresponding early warning prompt text and present it visually, so as to prompt the back-end management personnel to take corresponding measures through the early warning prompt text; When an appliance control signal is received, dynamic control operations are performed on the corresponding high-risk home appliances, and a corresponding dynamic control plan is set for the home appliance Q based on the user's behavior habits.

Images