| CPC H02J 3/003 (2020.01) [H02J 2203/10 (2020.01)] | 7 Claims |
|
1. A method for optimized scheduling of an integrated energy system, wherein the integrated energy system comprises a wind energy, a solar energy, and a hydrogen energy, the method comprising:
by using a processor, respectively obtaining marginal distribution functions of a unit electricity use consumption resource, a unit hydrogen energy harvesting resource, a wind power, a photovoltaic power, and a load within a preset time period according to historical data; determining a first joint distribution function of the unit electricity use consumption resource and the unit hydrogen energy harvesting resource and a second joint distribution function of the wind power, the photovoltaic power, and the load based on the marginal distribution functions; then predicting a unit electricity use consumption resource, a unit hydrogen energy harvesting resource, a maximum wind power, a maximum photovoltaic power, and a load within the preset time period based on the first joint distribution function and the second joint distribution function, wherein electricity use refers to use of electricity outside the integrated energy system, the hydrogen energy harvesting resource is obtained through transporting the hydrogen energy outward, and the load refers to electricity use amount; estimating the marginal distribution functions of the unit electricity use consumption resource, the unit hydrogen energy harvesting resource, the wind power, the photovoltaic power, and the load by adopting a non-parametric kernel density estimation method; generating the first joint distribution function of the unit electricity use consumption resource and the unit hydrogen energy harvesting resource using a two-dimensional Frank Copula function; generating the second joint distribution function of the wind power, the photovoltaic power, and the load using a three-dimensional Frank Copula function, wherein Copula link functions C(·) used in the above processes of generating the joint distribution functions are respectively:
 where v1 and v2 are respectively the marginal distribution functions of the unit electricity use consumption resource and the unit hydrogen energy harvesting resource, and θ1 is a parameter of the two-dimensional Frank Copula function; and
 where u1, u2, and u3 respectively represent the marginal distribution functions of the wind power, the photovoltaic power, and the load, and θ2 is a parameter of the three-dimensional Frank Copula function; sampling the first joint distribution function and the second joint distribution function, and inversely transforming a sampling result and the two joint distribution functions to obtain the unit electricity use consumption resource, the unit hydrogen energy harvesting resource, the maximum wind power, the maximum photovoltaic power, and the load corresponding to each time period;
by using the processor, adjusting a predicted value of the load based on a dynamic change of the unit electricity use consumption resource within the preset time period to obtain a load value after performing a demand response based on the unit electricity use consumption resource to reduce a peak-valley difference of the load, wherein the preset time period comprises a plurality of time periods, and the load value based on a unit electricity use consumption resource response is determined through a following formula:
 where qi is a load after an i-th time period demand response, q0i is a predicted value of the load before the i-th time period demand response, i=1, 2, . . . , m, where m is a total number of periods of the preset time period, Δpj is an increment of the unit electricity use consumption resource before and after a j-th time period demand response, Δpj=pj−pj0, j=1, 2, . . . , m, pj0 is the unit electricity use consumption resource before the j-th time period demand response, pj is the unit electricity use consumption resource after the j-th time period demand response, which is a predicted unit electricity use consumption resource, and E is an elasticity matrix of the load and the unit electricity use consumption resource;
 Δqi is a load change before and after the demand response, Δqi=qi−q0i, eij is an elasticity coefficient, when j=i, eij is a self-elasticity coefficient, and when j≠i, eij is a cross-elasticity coefficient;
by using the processor, obtaining a first objective function of a system operating consumption resource and a second objective function of an environmental consumption resource based on predicted values of the unit electricity use consumption resource, the unit hydrogen energy harvesting resource, the maximum wind power, and the maximum photovoltaic power and based on the load value of the unit electricity use consumption resource response, wherein the first objective function comprises an electricity use consumption resource, a penalty consumption resource for not using the wind energy and the solar energy, and a negative number of a resource harvested by converting the wind energy and the solar energy into the hydrogen energy, the second objective function comprises a carbon emission consumption resource caused by electricity use, the first objective function is: F1=min (fbuy+fdis−fsell,H), where F1 is the system operating consumption resource, fbuy is the electricity use consumption resource of the system, fdis is the penalty consumption resource for not using the wind energy and the solar energy of the system, and fsell,H is the resource harvested by converting the wind energy and the solar energy into the hydrogen energy;
 where βt is the predicted value of the unit electricity use consumption resource during a time period t and Pbuy (t) is the load during the time period t;
 where ω is a unit penalty consumption resource for not using the wind energy and the solar energy, and PW,dis (t) and PPV,dis (t) are respectively a wind abandonment power and a solar abandonment power during the time period t; the wind abandonment power refers to unused wind power, and the solar abandonment power refers to unused photovoltaic power;
 where αt is the unit hydrogen energy harvesting resource during the time period t and Hsell (t) is the hydrogen energy transported outward during the time period t; and
the second objective function is:
 where F2 is the environmental consumption resource, fCO2 is a tiered carbon emission consumption resource of the system, λ is a basic consumption resource of carbon emission, l is an interval division length of carbon emission, α is a growth rate of consumption resource, and EIESt is a carbon emission amount;
 where EIESa is an actual carbon emission amount generated by the system, EIES is a total carbon emission quota allocated to the system, χca is a carbon emission amount generated by the system after a unit of electricity is used, Pbuy (t) is the load, and χc is a free carbon emission quota obtained by the system after a unit of electricity is used from outside the system;
by using the processor, weighting and summing the first objective function and the second objective function to obtain an overall objective function, and then solving for the overall objective function to obtain an optimized scheduling result of the integrated energy system within the preset time period, so that an economic efficiency of the integrated energy system is improved and a carbon dioxide emission is reduced,
wherein the overall objective function is an optimized scheduling model established by using data generated by the two-dimensional Frank Copula function, the three-dimensional Frank Copula function and the load formula as input parameters,
wherein the optimized scheduling result comprises an electricity use, a battery output, a hydrogen energy storage output, a hydrogen energy export, a photovoltaic output and a wind power output during each time period,
wherein based on the optimized scheduling result, power is selectively apportioned to loads during peak demand periods, and the power includes the battery output, the hydrogen energy storage output, the hydrogen energy export, and wherein a battery regulates charge and discharge cycles, and regulates hydrogen production in response to the optimized scheduling result.
|