Multi-source micro-energy collection power distribution characteristic evaluation method and device and storage medium

Abstract

The application discloses a multi-source micro-energy collection power distribution characteristic evaluation method and device and a storage medium, and relates to the technical field of micro-energy collection. Firstly, a multi-natural environment variable related coupling Copula function model is established according to historical wind speed, illumination, temperature and radio field intensity information of an application site, then a sampling point needing to be evaluated is established based on the probability density distribution characteristics of the natural environment, then the total power corresponding to the output of a mixed micro-energy collection system of each sampling point is calculated based on the power output characteristic curves of each micro-energy collection device, further, the multi-source mixed micro-energy collection system output power evaluation function under the sampling point is established with each natural environment variable as the dependent variable, the sampling point depth is adaptively increased by nesting until the final output power evaluation function is determined, finally, an output power sample sequence is generated, and the multi-source mixed micro-energy collection system output power probability density function is determined based on kernel density estimation.

Description

Technical Field

[0001] This invention belongs to the field of micro-energy harvesting technology, and particularly relates to a method, device and storage medium for evaluating the power distribution characteristics of multi-source micro-energy harvesting. Background Technology

[0002] In recent years, micro-energy harvesting technology has become crucial for building green sensor networks. Single-source micro-energy harvesting devices suffer from unstable energy output due to environmental influences, while multi-source hybrid micro-energy harvesting devices, with their complementary characteristics among multiple energy sources, offer more stable energy output compared to single-source devices, making them an important direction for development in the field. Multi-source hybrid micro-energy harvesting devices comprise multiple energy harvesting units utilizing environmental micro-energy sources such as wind, light, temperature differences, and radio frequency. They rely on various mechanisms, including mechanical motion, photoelectric effect, thermoelectric power generation, vibration piezoelectricity, and electromagnetic induction, to achieve relatively stable energy output through the combined collection of multiple energy sources. Accurately assessing the probability distribution characteristics of usable energy in multi-source hybrid micro-energy harvesting devices is fundamental for energy management and evaluation analysis. However, due to the significant randomness and uncertainty inherent in wind, light, temperature, and radio frequency signals in the natural environment, the energy collected by multi-source hybrid micro-energy harvesting devices also exhibits randomness. Meanwhile, the power output characteristics of different types of energy harvesters vary greatly, and the evaluation of their system output power probability distribution characteristics is also more complex. Micro energy harvesting devices under different mechanisms also have very different power output characteristics and strong nonlinear characteristics. How to construct an accurate evaluation method for the system output power probability distribution characteristics is very important for the development of multi-source hybrid micro energy harvesting technology. Summary of the Invention

[0003] The purpose of this application is to provide a method, device, and storage medium for evaluating the power distribution characteristics of multi-source micro-energy harvesting.

[0004] To achieve the objectives of this application, the technical solution provided is as follows:

[0005] First aspect

[0006] This invention provides a method for evaluating the power distribution characteristics of multi-source micro-energy harvesting, comprising the following steps:

[0007] Step a: Based on historical wind speed, light intensity, temperature, and radio frequency field strength information of the application location of the micro-energy harvesting device, establish a multi-natural environmental variable coupled Copula function model:

[0008] This invention uses the R-vine Copula model to establish wind speed V w Irradiance r Temperature T c RF field strength E sThe joint probability density function of four-variable coupling. The R-tree used contains three trees, denoted as T1, T2, and T3. Each edge of each tree corresponds to a pair-copula, and the variables of the i-th tree constitute the node set N. i The pair-copula between nodes constitutes the edge set E. i Its structure is shown in the attached figure. Figure 2 As shown, the following conditions are met: 1) The node set N1 of tree T1 is {1,2,3,4}, and the edge set is E1; 2) The node set N2 of the second tree T2 is E1, and the edge set of two edges of tree T1 that share a common node is connected by an edge in tree T2, with the edge set being E2; 3) The node set N3 of the third tree T3 is E2. (See diagram c) i,j c i,j (u i ,u j (i,j=1,2,...,4,i≠j), c(·) denotes the Copula probability density function, c i,j|m c i,j|m (F(u i |u m ),F(u j |u m )), c i,j|m,n 。 indicates c i,j|m,n (F(u i |u m ,u n ),F(u j |u m ,u n F, the conditional distribution function, can be expressed by the following formula:

[0009]

[0010] The edge of the first tree is at wind speed V w Irradiance r Temperature T c RF field strength E sThe first tree uses three two-dimensional Copula functions to establish variables based on the distribution function. The second tree's edges use two two-dimensional Copula functions to establish variables based on conditional distributions F(u1|u2), F(u2|u3), and F(u3|u4). The third tree only needs one two-dimensional Copula function, which is the desired R-vine model. The two-dimensional Copula functions described in this invention include normal-copula, t-copula, Clayton-copula, Frank-copula, Gumbel-copula, and R-vine Pair-copula based on two-dimensional Copula functions. When applying these functions, the one with the smallest Euclidean distance from the empirical Copula function is selected as the final Copula function.

[0011] (2) Based on the probability density distribution characteristics of the natural environment, the nested sparse grid method based on Wasserstein distance is used to determine the sampling points that need to be evaluated.

[0012] The formula for selecting points using the sparse mesh method is:

[0013]

[0014] In the formula, q = k + d - 1, q is the collocation depth, k is the Smolyak level of the multi-indicator system, d is the dimension of the variable, and i = (i1, i2 = i d The set of coordinate indices for each variable, |i| = i1 + i2 + ... + i d A set of sampling points at a depth of q. It can be represented as In the formula Let r be the set of sampling points for the r-th variable.

[0015] From the expression A(q,d), we can see that:

[0016]

[0017] Where, Δ i =u i -u i-1 The difference between the selection of points at depth level i and depth level i-1.

[0018] Define e i Let e ​​be the error between the power obtained at the (i-1)th level depth sampling point and the power at the newly added sampling point at the i-th level. i =f(Δ i )-A(i-1,d)f(Δ i If the error is e, then the error is e. i The infinite norm |e i | ∞This can be used as an indicator of the accuracy of output power evaluation at the i-th level depth sampling point; the smaller the value, the more accurate the evaluation. Therefore, |e i | ∞ As an indicator, the depth is increased step by step to determine the appropriate depth level for the sampling points. This invention uses a nested sparse grid method based on Wasserstein distance to adaptively determine the sampling points to be evaluated. The steps are as follows:

[0019] 1) Based on Wasserstein distance Centered on the highest point of the joint probability density distribution function of the R-copula model, each variable l is determined in the direction of each variable with a distance interval D. max gradation point values,

[0020] 2) Take k=1, i=(1,1,1…1),u 1 Then, corresponding to the center point, the MLS-HDMR method is used to calculate... f(x) The output power expression is obtained. P1(x)

[0021] 3) Let k = k + 1, and determine the difference Δ between the selected points at the k-th depth and the (k-1)-th depth using the sparse grid method. k Based on the power expression P k-1 (x) Determine each Δ k The output power corresponding to point P k-1 (Δ k The error value e) k And calculate the error index |e k | ∞

[0022] 4) Compare |e k | ∞ With the set threshold e th If so, then stop and move P. k (x) is used as the output power expression; otherwise, repeat step 3) until the error index meets the standard requirements or k reaches l. max The end, in the final Pk(x) As an expression for output power.

[0023] (3) Based on the power output characteristic curves of each micro energy harvesting device, calculate the total output power of the hybrid micro energy harvesting system at each sampling point. The steps are as follows: According to the power output characteristic curves, determine the corresponding power output at each sampling point, and add the power to obtain the total output power of the hybrid micro energy harvesting device under the operating conditions at each sampling point.

[0024] (4) Calculate the probability distribution characteristics of the device output power using the RS-HDMR method, and then establish the probability distribution characteristic curve of the output power to assess the risk of insufficient available energy. The steps are as follows:

[0025] The HDMR model is used to establish the relationship between the power output of a multi-source hybrid micro-energy harvesting system and the input from the natural environment, with wind speed V... w Irradiance r Temperature T c RF field strength E s The four variables are used as the input vector x = (V w ,I r ,T c E s The total output power of the corresponding device is denoted as f(x). HDMR models the output quantity f(x) as the sum of the independent actions and interactions of the input variables, and its mathematical expression is:

[0026]

[0027] In the formula, f0 is the zeroth-order component, representing the constant term. i (x i ) is a first-order component function, representing a single variable x. i Its effect on the output. ij (x i ,x j ) is a second-order component function, representing the variable x. i x j The effect of mutual coupling on the output, f ijk (x i ,x j ,x k f is a third-order component function. ijkl (x i ,x j ,x k ,x l The component function is a fourth-order component. The solution is obtained using the Cut-HDMR form, based on the sampling points r = {V}. wr ,I rr ,T cr E sr In the formula, each order component can be expressed as: zero-order component: f0 = f(r), first-order component: f i (x i )=f(x i )-f0, second-order component: f ij (x i ,x j )=f(x i ,x j )-f i (xi )-f j (x j -f0, third-order component: f ijk (x i ,x j ,x k )=f(x i ,x j ,x k )-f ij (x i ,x j )-f ik (x i ,x k )-f jk (x j ,x k )-f i (x i )-f j (x j )-f k (x k -f0, fourth-order component:

[0028] Where: f i (x i )=ξ(r1,r2,...,r i-1 ,x i ,r i+1 ,...,r n ) indicates that the i-th variable takes the value x. i The remaining variables are the system output power values ​​when they are taken as reference values. ij (x i ,x j )=f(r1,r2,...,r i-1 ,x i ,r i+1 ,...,x j ,r j+1 ,...,r n ) indicates that the i-th and j-th variables take x values ​​respectively. i and x j The system output power is the value when the other variables are taken as reference values. And so on.

[0029] To solve the Cut-HDMR model, this invention uses the MLS method to construct the component functions in the HDMR expression. It assumes that the function values ​​y(x) of the original function y(x) at n nodes in the solution domain Ω are known. i (i = 1, 2, ..., n), define f as the approximation function of y(x) in the solution domain Ω. h (x):

[0030]

[0031] Where, p i (x) is a basis function, a i (x) represents the coefficients of the basis functions, and m is the number of basis functions. A one-dimensional linear basis function can take the form p(x) = [1, x]. T The one-dimensional nonlinear basis function can be taken as p(x) = [1, x, x]. 2 ,x 3 ,...,x s ] T Two-dimensional basis functions can take the form p(x) i ,x j )=[1,x i ,x j ,x i x j ] T When the basis functions do not meet the accuracy requirements, the number of sampling points and the order of the basis functions can be increased. The deviation at the sampling point is minimized when the weighted square of the deviation between the approximate and actual values ​​is minimized; the corresponding weighted residual equation can be expressed as:

[0032]

[0033] In the formula, ω(x) i ) is node x i The weight function should satisfy non-negativity, compact support, monotonically decreasing and smoothness. A fourth-order spline weight function is adopted.

[0034]

[0035] In the formula, s ml This represents the influence range of the sample points.

[0036] Differentiating the weighted residual equation with respect to a(x) and setting the derivative to zero, we have: a(x) = A -1 (x)B(x)f(x). Where A(x)=P T W(x)P, B(x)=P T W(x), f=[f(x1),f(x2),...,f(x4)] T ,

[0037]

[0038] Substitute it into f h (x), then the fitting function f h The expression for (x) is:

[0039] f h (x)=p T (x)a(x)=pT (x)A -1 (x)B(x),

[0040] 5) f h (x) can be substituted to establish the relationship between the power output and the natural environment input. The coupled Copula function of the natural environment variables is sampled by the Monte Carlo method. The sampled points are substituted into the obtained functional relationship, and the total output power of the device is statistically analyzed. The probability distribution characteristics of its output power are obtained by using the kernel density estimation method.

[0041] Second aspect

[0042] Corresponding to the above method, the present invention provides a device for evaluating the power distribution characteristics of multi-source micro-energy harvesting, comprising the following units:

[0043] The Copula function model building unit is used to establish a Copula function model with multiple natural environmental variables based on historical wind speed, light intensity, temperature and radio frequency field strength information of the application location of the micro energy harvesting device.

[0044] The sampling point establishment unit is used to establish the sampling points to be evaluated based on the probability density distribution characteristics of the natural environment using a nested sparse grid method based on Wasserstein distance.

[0045] The total power calculation unit is used to calculate the total output power of the hybrid micro-energy harvesting system at each sampling point based on the power output characteristic curve of the micro-energy harvesting device.

[0046] The output power evaluation function establishment unit is used to establish the output power evaluation function of the multi-source hybrid micro-energy harvesting system at this sampling point using the RS-HDMR method with each natural environmental variable as the dependent variable.

[0047] The final output power evaluation function acquisition unit is used to establish the sampling point unit, the total power calculation unit, and the output power evaluation function establishment unit. It adaptively increases the depth of the sampling points by adding the output power error corresponding to the new sampling points until the final output power evaluation function is established.

[0048] The output power probability distribution characteristic acquisition unit is used to generate a sample sequence of related coupled Copula function models based on the Monte Carlo method, input the final output power evaluation function to generate the corresponding output power sample sequence, and determine the output power probability density function of the multi-source hybrid micro-energy harvesting system based on kernel density estimation to obtain the output power probability distribution characteristics.

[0049] Third aspect

[0050] Corresponding to the above method, the present invention provides a storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the above-described method for evaluating the power distribution characteristics of multi-source micro-energy harvesting.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] 1. This invention can accurately evaluate the power output distribution characteristics of multi-source hybrid micro-energy harvesting devices. Considering the significant randomness of wind, light, temperature, and radio frequency signals in the natural environment, as well as the differences in power output characteristics among various types of energy harvesters, an effective method for evaluating the output power probability density distribution characteristics has been established. This is of great importance to the development and application of multi-source hybrid micro-energy harvesting technology.

[0053] 2. The method proposed in this invention can reduce the evaluation and testing time, effectively solve the nonlinearity and uncertainty problems encountered in the evaluation of power output distribution of multi-source hybrid micro-energy harvesting devices, provide guidance for the practical application of multi-source hybrid micro-energy harvesting devices in different locations and seasons, and provide basic data support for the efficient regulation of multi-source hybrid micro-energy harvesting devices. Attached Figure Description

[0054] Figure 1 This is a flowchart of the multi-source micro-energy harvesting power distribution characteristic evaluation method of this application;

[0055] Figure 2 This is a structural diagram of the R-vine in the embodiments of this application. Detailed Implementation

[0056] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0057] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0058] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when used in this specification, the words “comprising” and / or “including” indicate the presence of features, steps, operations, parts or modules, components and / or combinations thereof.

[0059] like Figure 1As shown in the figure, this application provides a method for evaluating the power distribution characteristics of multi-source micro-energy harvesting, characterized by the following steps:

[0060] Step a: Based on the historical wind speed, light intensity, temperature and radio frequency field strength information of the application location of the micro-energy harvesting device, establish a multi-natural environmental variable related coupled Copula function model;

[0061] This step includes:

[0062] aa: After obtaining historical information on wind speed, light intensity, temperature, and radio frequency field strength in the installation area of ​​the multi-source hybrid micro-energy harvesting system, the wind speed V at the assessment location for each month from January to December will be evaluated. w Irradiance r Temperature T c RF field strength E s The data are summarized separately to generate a dataset of four elements for each month: wind speed, solar irradiance, temperature, and radio frequency field strength at the installation location.

[0063] ab: Using the R-Teng Copula model, wind speed V for each month is established based on the dataset. w Irradiance r Temperature T c RF field strength E s The joint probability density function of four variables is used, where Rtree contains three trees, denoted as T1, T2, and T3. Each edge of each tree corresponds to a pair-copula function, and the variables of the i-th tree constitute the node set N. i The pair-copula functions between nodes form the edge set E. i ;

[0064] ac: R-vine structure as Figure 2 As shown, where: 1) The node set N1 of tree T1 is {1,2,3,4}, and the edge set is E1; 2) The node set N2 of the second tree T2 is E1, and the edge set E2 is formed by connecting two edges of tree T1 with a common node; 3) The node set N3 of the third tree T3 is E2. (See diagram c) i,j c i,j (u i ,u j (i,j=1,2,...,4,i≠j), u1,u2,u3,u4 correspond to wind speeds V respectively. w Irradiance r Temperature T c RF field strength E s Four variables, c(·) represents the Copula probability density function, c i,j|m c i,j|m (F(u i |um ),F(u j |u m )), c i,j|m,n 。 indicates c i,j|m,n (F(u i |u m ,u n ),F(u j |u m ,u n F is the conditional distribution function, calculated using the following formula:

[0065]

[0066] ad: The edges of each tree in the R vine satisfy the following condition: the edge of the first tree T1 is at wind speed V w Irradiance r Temperature T c RF field strength E s The distribution function is used as the variable to establish three two-dimensional Copula functions. The edges of the second tree T2 are established with two two-dimensional Copula functions with the conditional distributions F(u1|u2), F(u2|u3), and F(u3|u4) as variables. The third tree only needs to establish one two-dimensional Copula function. This two-dimensional Copula is the desired R-vine model.

[0067] ae: The two-dimensional Copula function mentioned therein includes, but is not limited to, normal-copula, t-copula, Clayton-copula, Frank-copula, Gumbel-copula, and R-Teng Pair-copula based on two-dimensional Copula function. Each two-dimensional function is selected to build an R-Teng model, and the model is compared with the empirical Copula function. The function with the smallest Euclidean distance between the two is selected as the final Copula function used.

[0068] Step b: Based on the probability density distribution characteristics of the natural environment, the sampling points to be evaluated are established using the nested sparse grid method based on Wasserstein distance;

[0069] Step b includes:

[0070] b-1: Centered on the highest point of the joint probability density distribution function of the R-vine Copula model, at wind speed V w Irradiance r Temperature T c RF field strength E s Each variable l is determined along a distance interval D in each of the four directions. max The gradation point value, the Wasserstein distance between any two points x and y in the i-th direction is calculated using the formula: [formula missing]. u1, u2, u3, u4 correspond to wind speed V respectively. w Irradiance r Temperature T c RF field strength E s Four variables, F(u) i |·) is u i The set of marginal probability density distribution functions, where γ is one of the probability density functions, inf denotes the infimum, and E γ(x,y) [||xy||] is the expected value of the distance between x and y calculated according to the probability density function γ;

[0071] b-2: The set of collocation points established based on the sparse grid method and the optimal set of collocation points established in b-1. The optimal set of collocation points is related to the depth q of the collocation points, and has the optimal set of collocation points at depth q. in Let q be the set of sampling points for the r-th variable, q be the collocation depth, q = k + 3, k be the Smolyak level of the multi-indicator system, and i = (i1, i2, ..., i4) be the set of collocation indicators for each variable, |i| = i1 + i2 + ... + i d × represents the Cadir product;

[0072] b-3: The optimal set of depth collocation points H at level q q Corresponding selection formula u1, u2, u3, u4 correspond to wind speed V respectively. w Irradiance r Temperature T c RF field strength E s Four variables, For tensor product, Let m be the number of combinations of choosing m elements from n distinct elements, satisfying The difference set between the i-th level depth and the (i-1)-th level depth of the selected points for the r-th variable can be improved by using a nested approach of progressively increasing the Smolyak level k of multiple indicators.

[0073] Step c: Based on the power output characteristic curves of each micro-energy harvesting device, calculate the total output power of the hybrid micro-energy harvesting system at each sampling point;

[0074] This step includes:

[0075] A preferred embodiment of the device power output curve is to determine the power output of each device at each operating point in the k-level matching point set based on the output characteristic curve of each power device in the multi-source hybrid micro-energy harvesting system, and add the power of each device together to obtain the total output power of the hybrid micro-energy harvesting device under the operating conditions of each matching point. The device power output characteristic curve is the power characteristic curve provided by the manufacturer, or it can be obtained by laboratory testing.

[0076] Another preferred approach is to simulate the environment corresponding to each operating point in the k-level set of test points in a laboratory testing device, and directly measure the total output power of the multi-source hybrid micro-energy harvesting system using a power measurement device to obtain the total output power of the hybrid micro-energy harvesting system at each sampling point.

[0077] Step d: Using each environmental variable as the dependent variable, evaluate the output power of the multi-source hybrid micro-energy harvesting system at this sampling point using the RS-HDMR method;

[0078] This step includes:

[0079] d-1: Wind speed V w Irradiance r Temperature T c RF field strength E s The four variables are used as the input vector x = (V w ,I r ,T c E s The total output power of the corresponding device is denoted as f(x). The HDMR model is used to establish the relationship between the power output of the multi-source hybrid micro-energy harvesting system and the input from the natural environment. HDMR models the output f(x) as the sum of the independent actions and interactions of the input variables, and its mathematical expression is:

[0080]

[0081] In the formula, f0 is the zeroth-order component; f represents the constant term. i (x i ) is a first-order component function, representing a single variable x. i Effect on output; f ij (x i ,x j ) is a second-order component function, representing the variable x. i x j The effect of mutual coupling on the output; f ijk (x i ,x j ,x k f is a third-order component function. ijkl (x i ,xj ,x k ,x l ) is a fourth-order component function, representing the interaction between multiple variables.

[0082] d-2: The solution uses the Cut-HDMR form, where any sampling point r = {V} in the k-level collocation basis. wr ,I rr ,T cr E sr The corresponding Cut-HDMR expression has the following components: zero-order component f0 = f(r), first-order component f... i (x i )=f(x i )-f0, second-order component f ij (x i ,x j )=f(x i ,x j )-f i (x i )-f j (x j )-f0, third-order component f ijk (x i ,x j ,x k )=f(x i ,x j ,x k )-f ij (x i ,x j )-f ik (x i ,x k )-f jk (x j ,x k )-f i (x i )-f j (x j )-f k (x k -f0, fourth-order component: Where: f i (x i )=ξ(r1,r2,...,r i-1 ,x i ,r i+1 ,...,r n ) indicates that the i-th variable takes the value x. i When the other variables take reference values, the system output power, f ij (x i ,x j )=f(r1,r2,...,ri-1 ,x i ,r i+1 ,...,x j ,r j+1 ,...,r n ) indicates that the i-th and j-th variables take x values ​​respectively. i and x j The system output power is when the other variables are taken as reference values.

[0083] d-3: Solving the Cut-HDMR model using the moving least squares method, the Cut-HDMR approximation model in the solution domain can be written as Where p i (x) is a basis function, a i (x) represents the coefficients of the basis functions, and m is the number of basis functions. A one-dimensional linear basis function can take the form p(x) = [1, x]. T The one-dimensional nonlinear basis function can be taken as p(x) = [1, x, x]. 2 ,x 3 ,...,x s ] T Two-dimensional basis functions can take the form p(x) i ,x j )=[1,x i ,x j ,x i x j ] T When the basis functions do not meet the accuracy requirements, the depth of the collocation set and the order of the basis functions can be increased. When the deviation between the approximate value and the actual value is minimized, the corresponding weighted residual equation can be expressed as: Where ω(x) i ) is node x i The weighting function uses a fourth-order spline weighting function. in s ml To determine the influence range of the sample points, differentiate the weighted residual equation with respect to a(x) and set the derivative to zero. Then we have: a(x) = A -1 (x)B(x)f(x), where A(x)=P T W(x)P, B(x)=P T W(x), f(x)=[f(x1),f(x2),...,f(x4)] T ,

[0084]

[0085] Substitute it into f h (x) gives the power output function P that approximates HDMR. k (x) is: P k (x)=p(x)a(x).

[0086] Step e: Repeat bd, adaptively increasing the depth of sampling points by the output power error corresponding to the newly added sampling points, until the final output power evaluation function is established;

[0087] Let k = k + 1, and determine the difference Δ between the selected points at the k-th depth and the (k-1)-th depth using the sparse grid method. k Based on the power expression P k-1 (x) Determine each Δ k The output power corresponding to point P k-1 (Δ k The error value e) k And calculate the error index |e k | ∞ Comparison |e k | ∞ With the set threshold e th If the value is less than the threshold, then stop and set P. k (x) is used as the output power evaluation function; otherwise, step bd is repeated until the error index meets the standard requirements or k reaches l. max End, with the final P k (x) serves as the output power evaluation function.

[0088] Step f: Generate a sequence of related coupled Copula function models based on the Monte Carlo method, input them into the final output power evaluation function to generate the corresponding output power sample sequence, and determine the output power probability density function of the multi-source hybrid micro-energy harvesting system based on kernel density estimation.

[0089] In addition, corresponding to the above method embodiments, this invention also provides a multi-source micro-energy harvesting power distribution characteristic evaluation device, comprising the following units:

[0090] The Copula function model building unit is used to establish a Copula function model with multiple natural environmental variables based on historical wind speed, light intensity, temperature and radio frequency field strength information of the application location of the micro energy harvesting device.

[0091] The sampling point establishment unit is used to establish the sampling points to be evaluated based on the probability density distribution characteristics of the natural environment using a nested sparse grid method based on Wasserstein distance.

[0092] The total power calculation unit is used to calculate the total output power of the hybrid micro-energy harvesting system at each sampling point based on the power output characteristic curve of the micro-energy harvesting device.

[0093] The output power evaluation function establishment unit is used to establish the output power evaluation function of the multi-source hybrid micro-energy harvesting system at this sampling point using the RS-HDMR method with each natural environmental variable as the dependent variable.

[0094] The final output power evaluation function acquisition unit is used to establish the sampling point unit, the total power calculation unit, and the output power evaluation function establishment unit. It adaptively increases the depth of the sampling points by adding the output power error corresponding to the new sampling points until the final output power evaluation function is established.

[0095] The output power probability distribution characteristic acquisition unit is used to generate a sample sequence of related coupled Copula function models based on the Monte Carlo method, input the final output power evaluation function to generate the corresponding output power sample sequence, and determine the output power probability density function of the multi-source hybrid micro-energy harvesting system based on kernel density estimation to obtain the output power probability distribution characteristics.

[0096] The Copula function model building unit is used to establish a multi-natural environmental variable coupled Copula function model based on historical wind speed, light intensity, temperature, and radio frequency field strength information of the micro-energy harvesting device application location. Specifically, it includes:

[0097] a-1: After acquiring historical information on wind speed, light intensity, temperature, and radio frequency field strength in the installation area of ​​the multi-source hybrid micro-energy harvesting system, the evaluation location will be assessed according to the wind speed V at a preset cycle. w Irradiance r Temperature T c RF field strength E s The wind speed V at each installation location for each period is summarized separately. w Irradiance r Temperature T c RF field strength E s The four elements of the dataset;

[0098] a-2: Using the R-Teng Copula model, wind speed V for each month is established based on the aforementioned four-element dataset. w Irradiance r Temperature T c RF field strength E s The joint probability density function of four variables coupled together; where Rtree contains 3 trees, denoted as T1, T2, and T3, each edge of each tree corresponds to a pair-copula function, and the variables of the i-th tree constitute the node set N. i The pair-copula functions between nodes form the edge set E. i The edges of each tree in vine R satisfy the following condition: the edge of the first tree T1 is at a wind speed V. w Irradiancer Temperature T c RF field strength E s The distribution function is used as the variable to establish three two-dimensional Copula functions. The edges of the second tree T2 are established with two two-dimensional Copula functions with the conditional distributions F(u1|u2), F(u2|u3), and F(u3|u4) as variables. The third tree only needs to establish one two-dimensional Copula function. The two-dimensional Copula function established by the third tree is the R-vine model.

[0099] a-3: Select multiple types of Pair-copula functions to establish R-vine models. Compare the established R-vine models with empirical Copula functions and select the Copula function with the smallest Euclidean distance between them as the final Copula function to be used.

[0100] The sampling point establishment unit is used to establish the sampling points to be evaluated based on the nested sparse grid method of Wasserstein distance, specifically including:

[0101] b-1: Centered on the highest point of the joint probability density distribution function of the R-vine Copula model, at wind speed V w Irradiance r Temperature T c RF field strength E s Each variable l is determined along a distance interval D in each of the four directions. max The gradation point value, the Wasserstein distance between any two points x and y in the i-th direction is calculated using the formula: distance u1, u2, u3, u4 correspond to wind speed V respectively. w Irradiance r Temperature T c RF field strength E s ,F(u i |·) is u i The set of marginal probability density distribution functions, where γ is one of the probability density functions, inf denotes the infimum, and E γ(x,y) [||xy||] is the expected distance between x and y calculated according to the probability density function γ;

[0102] b-2: The set of collocation points established in step b-1 based on the sparse mesh method. The selected set of collocation points is related to the depth q of the collocation points, and has the optimal set of collocation points at depth q. in Let be the set of sampling points for the r-th variable, q be the collocation depth, q = k + 3, k be the Smolyak level of the multi-indicator system, and i = (i1, i2, ..., i4) be the set of collocation indicators for each variable, |i| = i1 + i2 + ... + i d × represents the Cadir product;

[0103] b-3: The optimal set of depth collocation points H at level q q Corresponding selection formula u1, u2, u3, u4 correspond to wind speed V respectively. w Irradiance r Temperature T c RF field strength E s , For tensor product, Let m be the number of combinations of choosing m elements from n distinct elements, satisfying The difference set between the i-th level depth and the (i-1)-th level depth of the selected points for the r-th variable is used. The depth of the preferred matching point set is increased by nesting multiple indicators, Smolyak level k, and the preferred matching point set is used as the sampling point.

[0104] Corresponding to the above method, the present invention also provides a storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the above-described method for evaluating the power distribution characteristics of multi-source micro-energy harvesting.

[0105] It should be noted that those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0106] It should be noted that any technical solutions not detailed in this application employ publicly known technologies.

[0107] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for evaluating the power distribution characteristics of multi-source micro-energy harvesting, characterized in that, Includes the following steps: Step a: Based on the historical wind speed, light intensity, temperature and radio frequency field strength information of the application location of the micro-energy harvesting device, establish a correlation-coupled Copula function model of multiple natural environmental variables; Step b: Based on the probability density distribution characteristics of the natural environment, the sampling points to be evaluated are established using the nested sparse grid method based on Wasserstein distance; Step c: Based on the power output characteristic curve of the micro-energy harvesting device, calculate the total output power of the hybrid micro-energy harvesting system at each sampling point; Step d: Using each natural environmental variable as the dependent variable, establish the output power evaluation function of the multi-source hybrid micro-energy harvesting system at the sampling point using the RS-HDMR method; Step e: Repeat steps bd, adaptively increasing the depth of sampling points by adding the output power error corresponding to the new sampling points, until the final output power evaluation function is established; Step f: Generate a sample sequence of the related coupled Copula function model based on the Monte Carlo method, input it into the final output power evaluation function to generate the corresponding output power sample sequence, and determine the output power probability density function of the multi-source hybrid micro-energy harvesting system based on the kernel density estimation method, and determine the output power probability distribution characteristics of the multi-source hybrid micro-energy harvesting system based on the output power probability density function. Step a, which establishes a multi-natural-environment-variable-correlated coupled Copula function model based on historical wind speed, light intensity, temperature, and radio frequency field strength information of the micro-energy harvesting device application location, includes: a-1: After obtaining historical information on wind speed, light intensity, temperature, and radio frequency field strength in the installation area of ​​the multi-source hybrid micro-energy harvesting system, the evaluation location will be assessed according to the wind speed at a preset cycle. Irradiance ,temperature RF field strength The data are summarized separately to generate the wind speed at the installation location for each period. Irradiance ,temperature RF field strength The four elements of the dataset; a-2: Using the R-Teng Copula model, wind speeds for each month were established based on the aforementioned four-element dataset. Irradiance ,temperature RF field strength The joint probability density function of the four-variable coupling; where Rt contains 3 trees, denoted as... , , Each edge of each tree corresponds to a pair-copula function, and the variables of the i-th tree constitute the node set. The pair-copula functions between nodes form an edge set. The edges of each tree in vine R satisfy the following conditions: the first tree The edge is determined by wind speed Irradiance ,temperature RF field strength The distribution function is established using three two-dimensional Copula functions for the variables, and the second tree. The edges are conditionally distributed , , Two two-dimensional Copula functions are established for the variables, and a two-dimensional Copula function is established for the third tree. The two-dimensional Copula function established for the third tree is determined as the R-vine model. a-3: Select multiple types of Pair-copula functions to establish R-vine models. Compare the established R-vine models with empirical Copula functions and select the Copula function with the smallest Euclidean distance between them as the final Copula function to be used.

2. The method for evaluating the power distribution characteristics of multi-source micro-energy harvesting according to claim 1, characterized in that: Step b uses a nested sparse grid method based on Wasserstein distance to establish the sampling points to be evaluated, including: b-1: Centered on the highest point of the joint probability density distribution function of the R-vine Copula model, at the wind speed Irradiance ,temperature RF field strength In each of the four directions, at intervals of distance Determine each variable above gradation point value, the first i The Wasserstein distance between any two points x and y in a direction is calculated using the formula: , Corresponding to wind speeds Irradiance ,temperature RF field strength , for The set of marginal probability density distribution functions, Let be one of the probability density functions, and let inf denote the infimum. For probability density function Calculate the expected distance between x and y; b-2: Select the collocation set established in step b-1 based on the sparse mesh method. The selected collocation set is related to the collocation depth. q Related to, having the first q Deep set of points ,in For the first r The set of sampling points for each variable, where q is the depth of the sampling points, q=k+3, and k is the number of indicators; In step c, the power output characteristic curve of the micro-energy harvesting device is classified as Smolyak level. This is a set of indicators for the allocation of points for each variable. , For the accumulation of Kadir; b-3: Set of collocations at depth q-th order Corresponding selection formula , Corresponding to wind speeds Irradiance ,temperature RF field strength , For tensor product, Let m be the number of combinations of choosing m elements from n distinct elements, satisfying , The difference set between the i-th level depth and the (i-1)-th level depth of the selected points is used. The depth of the collocation set is increased in a nested manner by gradually increasing the Smolyak level k of multiple indicators, and the collocation set is used as the sampling points to be evaluated.

3. The method for evaluating the power distribution characteristics of multi-source micro-energy harvesting according to claim 1, characterized in that, The power output characteristic curve of the micro-energy harvesting device in step c is determined in the following manner: To determine the power output of each device at each operating point in the k-level set of sampling points based on the output characteristic curves of each power device in the multi-source hybrid micro-energy harvesting system, the power output of each device is summed to obtain the total output power of the hybrid micro-energy harvesting device under the operating conditions of each sampling point. The power output characteristic curve is either the power characteristic curve configured for the device or obtained through laboratory testing. The environment corresponding to each operating point in the k-level set of sampling points is simulated in the laboratory testing equipment, and the total output power of the multi-source hybrid micro-energy harvesting system is measured by a power measurement device to obtain the total output power of the hybrid micro-energy harvesting system at each sampling point.

4. The method for evaluating the power distribution characteristics of multi-source micro-energy harvesting according to claim 1, characterized in that, Step d uses each environmental variable as the dependent variable and employs the RS-HDMR method to evaluate the output power of the multi-source hybrid micro-energy harvesting system at this sampling point, including: d-1: Wind speed Irradiance ,temperature RF field strength As the input vector x=( , , , The total output power of the corresponding multi-source hybrid micro-energy harvesting system is denoted as . The HDMR model is used to establish the relationship between the power output of a multi-source hybrid micro-energy harvesting system and the input from the natural environment. The HDMR model will output... The model is the sum of the independent actions and interactions of the input variables, and its mathematical expression is: ； In the formula, It is a zeroth-order component, represented as a constant term; It is a first-order component function, representing a single variable. Effect on output; It is a second-order component function, representing the variable , The effect of mutual coupling on the output; It is a third-order component function; It is a fourth-order component function, representing the interaction between multiple variables; d-2: The solution uses the Cut-HDMR form, with any sampling point in the k-level collocation set. The components in the corresponding Cut-HDMR expression are: zero-order component First-order components Second-order components , Third-order components Fourth-order components: ,in: Indicates the first Take one variable The system output power is when the other variables take reference values. Indicates the first , Each variable is taken and The system output power is when the other variables take reference values; d-3: Solving the Cut-HDMR model using the moving least squares method, the Cut-HDMR approximation model in the solution domain is expressed as follows: ,in These are basis functions. The coefficients of the basis functions, m It is the number of basis functions; one-dimensional linear basis functions are selected. One-dimensional nonlinear basis functions are taken Two-dimensional basis functions take When the basis functions do not meet the accuracy requirements, increasing the depth of the collocation set and the order of the basis functions minimizes the deviation between the approximate and actual values. The corresponding weighted residual equation is expressed as: ,in For nodes The weighting function uses a fourth-order spline weighting function. ,in , To determine the influence range of the sample points, the weighted residual equation is applied to... Taking the derivative and setting it to zero, we have: ,in, , , , ， ； Bring it in The power output function approximated by HDMR is obtained. for: .

5. The method for evaluating the power distribution characteristics of multi-source micro-energy harvesting according to claim 4, characterized in that, In step e, bd is repeated, and the depth of the sampling points is adaptively increased by the output power error corresponding to the newly added sampling points until the final output power evaluation function is established, including: make k is a natural number greater than 2. The difference between the selected points at the k-th depth and the (k-1)-th depth is determined using the sparse grid method. Based on power expression Determine each The output power corresponding to the point and error value And calculate the error index. ; Compare With set threshold ,like Less than the set threshold Then Use it as the output power evaluation function; otherwise, repeat step bd until the error index meets the standard requirements or k reaches the target. Finish, For the preset maximum level, with the final As an output power evaluation function.

6. A device for evaluating the power distribution characteristics of multi-source micro-energy harvesting, characterized in that, include: The Copula function model building unit is used to establish a Copula function model with multiple natural environmental variables based on historical wind speed, light intensity, temperature and radio frequency field strength information of the application location of the micro energy harvesting device. The sampling point establishment unit is used to establish the sampling points to be evaluated based on the probability density distribution characteristics of the natural environment using a nested sparse grid method based on Wasserstein distance. The total power calculation unit is used to calculate the total output power of the hybrid micro-energy harvesting system at each sampling point based on the power output characteristic curve of the micro-energy harvesting device. The output power evaluation function establishment unit is used to establish the output power evaluation function of the multi-source hybrid micro-energy harvesting system at the sampling point using the RS-HDMR method with each natural environmental variable as the dependent variable. The final output power evaluation function acquisition unit is used to establish the sampling point unit, the total power calculation unit, and the output power evaluation function establishment unit. It adaptively increases the depth of the sampling points by adding the output power error corresponding to the new sampling points until the final output power evaluation function is established. The output power probability distribution characteristic acquisition unit is used to generate a sample sequence of related coupled Copula function models based on the Monte Carlo method, input the final output power evaluation function to generate the corresponding output power sample sequence, and determine the output power probability density function of the multi-source hybrid micro-energy harvesting system based on kernel density estimation to obtain the output power probability distribution characteristics. The Copula function model building unit is used to establish a multi-natural environmental variable coupled Copula function model based on historical wind speed, light intensity, temperature, and radio frequency field strength information of the micro-energy harvesting device application location. Specifically, it includes: a-1: After obtaining historical information on wind speed, light intensity, temperature, and radio frequency field strength in the installation area of ​​the multi-source hybrid micro-energy harvesting system, the evaluation location will be assessed according to the wind speed at a preset cycle. Irradiance ,temperature RF field strength The data are summarized separately to generate the wind speed at the installation location for each period. Irradiance ,temperature RF field strength The four elements of the dataset; a-2: Using the R-Teng Copula model, wind speeds for each month were established based on the aforementioned four-element dataset. Irradiance ,temperature RF field strength The joint probability density function of the four-variable coupling; where Rt contains 3 trees, denoted as... , , Each edge of each tree corresponds to a pair-copula function, and the variables of the i-th tree constitute the node set. The pair-copula functions between nodes form an edge set. The edges of each tree in vine R satisfy the following conditions: the first tree The edge is determined by wind speed Irradiance ,temperature RF field strength The distribution function is established using three two-dimensional Copula functions for the variables, and the second tree. The edges are conditionally distributed , , Two two-dimensional Copula functions are established for the variables. Only one two-dimensional Copula function needs to be established for the third tree. The two-dimensional Copula function established for the third tree is the R-vine model. a-3: Select multiple types of Pair-copula functions to establish R-vine models. Compare the established R-vine models with empirical Copula functions and select the Copula function with the smallest Euclidean distance between them as the final Copula function to be used.

7. The multi-source micro-energy harvesting power distribution characteristic evaluation device according to claim 6, characterized in that: The sampling point establishment unit is used to establish the sampling points to be evaluated based on the nested sparse grid method of Wasserstein distance, specifically including: b-1: Centered on the highest point of the joint probability density distribution function of the R-vine Copula model, at the wind speed Irradiance ,temperature RF field strength In each of the four directions, at intervals of distance Determine each variable above gradation point value, the first i The Wasserstein distance between any two points x and y in a direction is calculated using the formula: , Corresponding to wind speeds Irradiance ,temperature RF field strength , for The set of marginal probability density distribution functions, Let be one of the probability density functions, and let inf denote the infimum. For probability density function Calculate the expected distance between x and y; b-2: Select the collocation set established in step b-1 based on the sparse mesh method. The selected collocation set is related to the collocation depth. q Related to, having the first q Deep set of points ,in For the first r The set of sampling points for each variable, where q is the depth of the collocation, q = k + 3, and k is the Smolyak level of the multi-indicator system. This is a set of indicators for the allocation of points for each variable. , For the accumulation of Kadir; b-3: Set of collocations at depth q-th order Corresponding selection formula , Corresponding to wind speeds Irradiance ,temperature RF field strength , For tensor product, Let m be the number of combinations of choosing m elements from n distinct elements, satisfying , The difference set between the i-th level depth and the (i-1)-th level depth of the selected points for the r-th variable is used to increase the depth of the collocation set in a nested manner by gradually increasing the Smolyak level k of multiple indicators, and the collocation set is used as the sampling point.

8. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method for evaluating the power distribution characteristics of multi-source micro-energy harvesting as described in any one of claims 1 to 5.

Images