MATLAB|【免费】高比例可再生能源电力系统的调峰成本量化与分摊模型

目录

主要内容

部分代码

结果一览

下载链接


主要内容

程序复现文献《高比例可再生能源电力系统的调峰成本量化与分摊模型》,从净负荷波动的角度出发,建立了调峰成本的量化与分摊模型,构造了无调峰需求的替代场景,将负荷和可再生能源出力曲线分别转换为无波动的均值线。其次,建立了含深度调峰和抽水蓄能的调度优化模型,用于计算不同场景下的调峰成本,并将有无调峰需求两种场景下的系统调峰成本之差作为单一主体导致的边际调峰成本,然后采用 Shapley值计算不同主体导致的调峰成本。最后,根据成本的引发程度分摊调峰成本。程序采用matlab+cplex编写,程序已经深度优化,求解速度很快,注释清晰,代码免费分享给大家学习参考!

部分代码

复制代码
for t=1:24
    C=[C, sum(sum(P300_Diot(:,:,t)))+sum(sum(P600_Diot(:,:,t)))+sum(P200_it(:,t))+P_PV(t)+P_Wind(t)+sum(P_PuG(:,t)-P_PuS(:,t))+P_guding == L_jun(t)+D_Eload_FW(t)+D_Eload_JM(t)+D_Eload_ZZ(t),
        P_Wind(t) == 0;     
        P_PV(t) == 0;    
        ];                       
end                                 
%%  公式(8-9)系统正负旋转备用约束
u200Git = binvar(20,24,'full');   %uit 表示机组运行状态的0-1变量,1开机,0关机
u300Git = binvar(18,24,'full');  
u600Git = binvar(18,24,'full');  
R_Ureq = (Eload*0.05+E_Wind*0.05+E_PV*0.05);    %正旋转备用
R_Dreq = (Eload*0.05+E_Wind*0.05+E_PV*0.05);    %负旋转备用
for t=1:24         
    C=[C,   200*sum(u200Git(:,t))+300*sum(u300Git(:,t))+600*sum(u600Git(:,t))-sum(sum(P300_Diot(:,:,t)))-sum(sum(P600_Diot(:,:,t)))-sum(P200_it(:,t)) >= R_Ureq(t),   %正旋转备用约束
            sum(sum(P300_Diot(:,:,t)))+sum(sum(P600_Diot(:,:,t)))+sum(P200_it(:,t))-200*0.5*sum(u200Git(:,t))-300*0.35*sum(u300Git(:,t))-600*0.35*sum(u600Git(:,t)) >= R_Dreq(t),  %负旋转备用约束     
         ];      
end   
%%  公式(10-12)火电机组功率大小约束,这里换一种方法去写
uP300_Diot = binvar(18,4,24,'full');
uP600_Diot = binvar(18,4,24,'full');
for t= 1:24   
    for i = 1:18 
            C = [C,  
                     sum(uP300_Diot(i,:,t))<= u300Git(i,t),  
                     sum(uP600_Diot(i,:,t))<= u600Git(i,t),  
                     u200Git(i,t)*200*0.5<=P200_it(i,t),P200_it(i,t)<=u200Git(i,t)*200, 
                     uP300_Diot(i,1,t)*300*0.5<=P300_Diot(i,1,t),P300_Diot(i,1,t)<=uP300_Diot(i,1,t)*300,
                     uP300_Diot(i,2,t)*300*0.45<=P300_Diot(i,2,t),P300_Diot(i,2,t)<=uP300_Diot(i,2,t)*300*0.5,
                     uP300_Diot(i,3,t)*300*0.4<=P300_Diot(i,3,t),P300_Diot(i,3,t)<=uP300_Diot(i,3,t)*300*0.45,
                     uP300_Diot(i,4,t)*300*0.35<=P300_Diot(i,4,t),P300_Diot(i,4,t)<=uP300_Diot(i,4,t)*300*0.4,
                     uP600_Diot(i,1,t)*600*0.5<=P600_Diot(i,1,t),P600_Diot(i,1,t)<=uP600_Diot(i,1,t)*600,
                     uP600_Diot(i,2,t)*600*0.45<=P600_Diot(i,2,t),P600_Diot(i,2,t)<=uP600_Diot(i,2,t)*600*0.5,
                     uP600_Diot(i,3,t)*600*0.4<=P600_Diot(i,3,t),P600_Diot(i,3,t)<=uP600_Diot(i,3,t)*600*0.45,
                     uP600_Diot(i,4,t)*600*0.35<=P600_Diot(i,4,t),P600_Diot(i,4,t)<=uP600_Diot(i,4,t)*600*0.4,
                     ];
    end
    for i = 19:20
            C = [C,  u200Git(i,t)*200*0.5<=P200_it(i,t),P200_it(i,t)<=u200Git(i,t)*200, 
                     ];
    end
end
%公式(11-12)不用再写了,已经隐含在上式(10)里了。
%%   公式(13-14)
%约束的含义是,持续运行需要受到上下爬坡约束,但启停不受爬坡约束。火电机组爬坡率为2%/min,小时调度体现不出来
for  i= 1:18
    for t = 2:24  
        C = [C,
               sum(P300_Diot(i,:,t))-sum(P300_Diot(i,:,t-1)) <= 0.3*300+(1-u300Git(i,t-1))*300,
               sum(P300_Diot(i,:,t-1))-sum(P300_Diot(i,:,t)) <= 0.3*300+(1-u300Git(i,t))*300,
               sum(P600_Diot(i,:,t))-sum(P600_Diot(i,:,t-1)) <= 0.3*600+(1-u600Git(i,t-1))*600,
               sum(P600_Diot(i,:,t-1))-sum(P600_Diot(i,:,t)) <= 0.3*600+(1-u600Git(i,t))*600,               
               P200_it(i,t)-P200_it(i,t-1) <= 0.3*200+(1-u200Git(i,t-1))*200,
               P200_it(i,t-1)-P200_it(i,t) <= 0.3*200+(1-u200Git(i,t))*200,
              ];
    end       
    for i = 19:20
            C = [C, P200_it(i,t)-P200_it(i,t-1) <= 0.3*200+(1-u200Git(i,t-1))*200,
                    P200_it(i,t-1)-P200_it(i,t) <= 0.3*200+(1-u200Git(i,t))*200,
                     ];
    end
end       
%%   公式(15-16)   火电机组的启停状态标识位置
C = [C,
      y200_it(:,2:24)-z200_it(:,2:24)==u200Git(:,2:24)-u200Git(:,1:23),  %公式(15)
      y300_it(:,2:24)-z300_it(:,2:24)==u300Git(:,2:24)-u300Git(:,1:23),  
      y600_it(:,2:24)-z600_it(:,2:24)==u600Git(:,2:24)-u600Git(:,1:23),  
      y200_it(:,1)==0,y300_it(:,1)==0,y600_it(:,1)==0,
      z200_it(:,1)==0,z300_it(:,1)==0,z600_it(:,1)==0,
      y200_it+z200_it<=1,%公式(16)
      y300_it+z300_it<=1,
      y600_it+z600_it<=1,
    ];   
%%   公式(17-18)   最短运行时间,最短停机时间限值,取20h和10h
%需要注意的是,调度的维度只有24,因此,编写代码的时候,需要避免调用越界24
Ton = 20;%最短开机时间          
Toff = 10;%最短关机时间          
for t=1:(25-Ton)   %公式(17)
    for i=1:18
         C = [C,sum(u200Git(i,t:(t+Ton-1))) >= Ton*y200_it(i,t)];
         C = [C,sum(u300Git(i,t:(t+Ton-1))) >= Ton*y300_it(i,t)];
         C = [C,sum(u600Git(i,t:(t+Ton-1))) >= Ton*y600_it(i,t)];
    end
    for i=19:20   
        C = [C,sum(u200Git(i,t:(t+Ton-1))) >=Ton*y200_it(i,t)];
    end
end
for t=1:(25-Toff)    %公式(18)
    for i=1:18   
         C = [C,sum(1-u200Git(i,t:(t+Toff-1))) >= Toff*z200_it(i,t),
              sum(1-u300Git(i,t:(t+Toff-1))) >= Toff*z300_it(i,t),
              sum(1-u600Git(i,t:(t+Toff-1))) >= Toff*z600_it(i,t)];
    end
    for i=19:20    
        C = [C,sum(1-u200Git(i,t:(t+Toff-1))) >=Toff*z200_it(i,t)];
    end  
end   
%%  公式(19)  可再生能源出力约束
C = [C, 0<=P_PV,P_PV<=E_PV,
        0<=P_PVcur,P_PVcur<=E_PV,
        P_PV+P_PVcur==E_PV,
        0<=P_Wind,P_Wind<=E_Wind,
        0<=P_Windcur,P_Windcur<=E_Wind,
        P_Wind+P_Windcur==E_Wind,
        ];    
%%  公式(20-24)  抽水蓄能电站功率约束
U_PuG = binvar(4,24,'full');
U_PuS = binvar(4,24,'full');
u_PuG = binvar(1,24,'full');
u_PuS = binvar(1,24,'full');
C = [C, 
         U_PuG*300*0.3<=P_PuG,P_PuG<=U_PuG*300,   %公式(20)
         U_PuS*300*0.3<=P_PuS,P_PuS<=U_PuS*300,   %公式(21)
         u_PuG+u_PuS<=1,   %公式(22)
        ];  
for t=1:24
    C = [C, 
         U_PuG(:,t)<=u_PuG(1,t),   %公式(23)
         U_PuS(:,t)<=u_PuS(1,t),   %公式(24)
        ];  
end    
​

结果一览

下载链接

相关推荐
Evand J5 小时前
LOS/NLOS环境建模与三维TOA定位,MATLAB仿真程序,可自定义锚点数量和轨迹点长度
开发语言·matlab
孤亭远见10 小时前
COMSOL with Matlab
matlab
图南楠12 小时前
simulink离散传递函数得到差分方程并用C语言实现
matlab
信号处理学渣12 小时前
matlab画图,选择性显示legend标签
开发语言·matlab
机器学习之心1 天前
Bayes-GRU-Attention的数据多特征分类预测Matlab实现
matlab·分类·gru
叶庭云1 天前
Matlab 和 R 语言的数组索引都是从 1 开始,并且是左闭右闭的
matlab·编程语言·r·数组索引·从 1 开始
γ..1 天前
基于MATLAB的图像增强
开发语言·深度学习·神经网络·学习·机器学习·matlab·音视频
IT猿手1 天前
基于PWLCM混沌映射的麋鹿群优化算法(Elk herd optimizer,EHO)的多无人机协同路径规划,MATLAB代码
算法·elk·机器学习·matlab·无人机·聚类·强化学习
超级大咸鱼2 天前
CW信号的正交解调
matlab·verilog·fpga·数字信号·解调·正交解调·cw
乌恩大侠2 天前
画图,matlab,
开发语言·matlab