Wind and light model output based on Monte Cano (Matlab code implementation)

Table of contents

1 Overview

2 Photovoltaic DG output probability model

2.1 Light intensity distribution characteristics

2.2 Photovoltaic DG active output

3 Wind power DG output probability model

3.1 Wind speed distribution characteristics

3.2 Active output of wind power DG

4 Matlab implementation

1 Overview

(1) Large-scale new energy crowds the channel, and the consumption capacity and utilization rate are mutually restricted.

Large-scale new energy is connected to the grid, the proportion of new energy power and electricity continues to increase, and the contradiction between consumption and utilization is prominent: the channel is crowded with conventional units, and the pressure on important sections to operate at full power is high; the anti-peaking characteristics of wind power are prominent, which intensifies the grid Peak shaving is difficult, and consumption and utilization are mutually restricted. On the basis of ensuring the consumption level, the utilization rate has a downward trend. If there is a 500kV UHV transmission line near a large wind farm, the wind power utilization rate of this wind farm can basically reach close to 100%, and it can be transmitted to distant places through the UHV line; if there is no UHV line, it can only be consumed nearby. Also, new energy can only produce active power. Due to power angle factors, sometimes the current does not flow in the direction we want, for example, some of it flows to other areas. It is necessary to install large reactive devices in some places to form a "section". , to prevent the flow of new energy.

(2) There are many uncontrollable factors in new energy output, and the uncertainty of power balance is large.

The output of wind and new energy is greatly affected by bad weather and climate, and short-term, medium- and long-term power supply is in a state of difficulty to predict. Affected by wind direction, wind force, light intensity, etc., the output of scenery has the characteristics of randomness, volatility, and intermittency, and the output of new energy sources fluctuates within a day; in extreme weather such as cold waves or continuous sunshine and high temperature, the demand for electricity increases sharply, but the output of new energy sources is minimal. It may be at a low level, and the balance support capacity is insufficient.

(3) The equivalent moment of inertia of the system decreases, and frequency regulation and stability problems are highlighted

The scale of new energy grid-connected and the proportion of small and medium-sized power grids continue to increase, conventional units are replaced by a large number of power electronic equipment, the system's rotational inertia continues to decrease, and the frequency regulation capability is severely reduced.

In August 2019, a major power outage occurred in London, England, with a loss of about 930,000 kilowatts of load, and about 1 million people were affected by the power outage. This is the largest power outage accident in the UK that has affected the largest population since the "London Blackout" in 2003! After analysis, the cause of the accident is that the installed capacity of wind power and new energy accounted for 34.71%, the equivalent moment of inertia of the system was insufficient, and the system frequency After the unit goes off-grid continuously, it drops steeply to 49Hz or less, and the low-frequency load shedding action cuts off 5% of the load to cause a power outage.

(4) Reactive power support capacity declines, and voltage control and stability problems are prominent

On the one hand, new energy voltage regulation capabilities are insufficient, and voltage control is difficult in large-scale new energy grid-connected areas. On the other hand, dynamic reactive power support capabilities in high-proportion power-receiving areas are insufficient, and system voltage regulation capabilities continue to decline. System security is facing the risk of voltage instability.

DC and new energy units do not have the dynamic reactive power support capability of conventional power sources.

As the speed increases, the dynamic reactive power reserve of the system drops sharply. UHV DC faults induce voltage stability problems, which manifest as overvoltage for the sending-end power grid; continuous low voltage or even voltage collapse for the receiving-end power grid, and the system faces the risk of large-scale power outages.

(5) Globalization of cross-area AC-DC coupling faults, and non-I-frequency stability problems

With the expansion of new energy grid-connected scale and the rapid development of UHV AC-DC Internet, under the impact of disturbance events, the energy absorption and discharge characteristics of the system are more complex and present a new form.

The transmission and receiving end power grids are closely coupled by AC and DC, and the integration characteristics of the power grid are remarkable. The dynamic interaction characteristics of AC and DC across regions are more complex, and local transient energy impacts can easily cause global stability problems. Large-scale grid-connection of new energy, deep adjustment of power structure and stable form of power grid are more complicated, and the fast response characteristics of power electronic devices cause non-I frequency stability problems.

2 Photovoltaic DG output probability model

2.1 Light intensity distribution characteristics

The active power produced by a photovoltaic generator usually depends on the light intensity at the location where the photovoltaic generator is located. In order to obtain more accurate probability distribution characteristics, it is very important to adopt an effective method to fit the measured light intensity data.

The above picture shows the measured light intensity data of a photovoltaic power station from 7:00 to 18:00 throughout the year. It can be seen from the figure that the wind speed does have certain distribution characteristics, so this paper uses Beta distribution and Normal distribution to simulate the light intensity data with the actual measurement combine.

(1) Beta distribution

Probability density function:

In the formula: $r$ is the light intensity $(W /m^{2})$ , $\get()$ is the Gamma function, $r_{max}$ is the maximum light intensity, $\alpha$ and $\beta$ is the parameter of the Beta distribution. The parameters are determined by fitting the actual data.

(2) Normal distribution

Probability density function:

In the formula: $r$ is the light intensity $(W/m^{2})$ , $\sigma$ and $\mu$ is the parameter of the Normal distribution. The parameters are determined by fitting the actual data.

2.2 Photovoltaic DG active output

Photovoltaic power = Beta distribution * maximum light intensity * total module area * photoelectric conversion rate

$P_{s}=f(\frac{r}{r_{max}})\cdot r_{max}\cdot S\cdot \eta$

% 光伏有功出力样本
        pv_samp(1,:)=betarnd(a_pv,b_pv,1,times);%生成形状为(1,times,)，参数为a_pv,b_pv的Beta分布
        Ppv_samp(1,:)=pv_samp(1,:)*rmax*S_pv*prey_pv;% 光伏功率=Beta分布*最大光强*组件总面积*光电转换率

3 Wind power DG output probability model

3.1 Wind speed distribution characteristics

The active power generated by the wind turbine and the wind speed at the location of the wind turbine. It is particularly important to use an effective method to fit the measured wind speed data to obtain a more accurate probability distribution characteristic.

The figure above shows the annual measured wind speed data of the wind power station. It can be seen from the figure that the wind speed has certain distribution characteristics: skewed distribution. In this paper, Weibull distribution, Rayleigh distribution and Gamma distribution are used to fit the measured wind speed data .

(1) Gamma distribution

Probability density function:

Where: $\not$ is the wind speed (m/s), $\get()$ is the Gamma function, and α and β are the parameters of the Gamma distribution. The parameters are determined by fitting the actual data.
This article uses the Gamma distribution, of course, it can also be replaced by the following two distributions.

wt_samp =gamrnd(c_wt,k_wt,1,times);  % 风速 产生服从gammer分布的样本 ，形状为（(1,times,)
%wt_samp =wblrnd(c_wt,k_wt,1,times);  % 风速 可以修改为服从weibull分布的样本 ，形状为（(1,times,)

(2) Rayleigh distribution

Probability density function:

In the formula: $\not$ is wind speed (m/s), $\sigma$ is the parameter of Rayleigh distribution. The parameters are determined by fitting the actual data.
The Rayleigh distribution is a special case of the Weibull distribution. It is usually suitable for long-term (one year) and above wind speed statistics. There is a 10% error between the fitting result and the actual result.

(3) Weibull distribution

Probability density function:

Where: $\not$ is the wind speed (m/s), $\lambda$ and , $k$ are the two parameters of Weibull distribution. The parameters are determined by fitting the actual data.
Weibull distribution is widely used, simple in form and simple in calculation, but its disadvantage is that it cannot adapt to the wind speed distribution of extreme weather in some areas.

3.2 Active output of wind power DG

After obtaining the wind speed model, it is necessary to obtain the formula of wind turbine output and wind speed:

Where: $\not$ is the wind speed, $\no _{but}$ is the cut-in wind speed, $\nu _{what}$ is the cut-out wind speed, $\no _{r}$ , is the rated wind speed of the wind turbine, and $P_{r}$ is the rated output power of the wind turbine.

4 Matlab implementation

 %% 风机出力
    for i=1:times  %得到风电出力样本
        if wt_samp(i)<vci %如果风速小于切入风速
            Pwt_samp(i)=0; %风机功率为0
        end 
        if wt_samp(i)>vci&&wt_samp(i)<vN %如果风速大于切入风速，同时小于额定风速
            Pwt_samp(i)=(wt_samp(i)-vci)/(vN-vci)*PN_wt;
            if   Pwt_samp(i)>PN_wt %如果风电功率大于额定功率
                Pwt_samp(i)=PN_wt; %则风电功率等于额定功率
            end 
        end
        if wt_samp(i)>vN&&wt_samp(i)< vco %如果风速大于额定风速 同时小于切出风速
            Pwt_samp(i)=PN_wt;%风电功率等于额定功率
        end 
        if wt_samp(i)> vco %如果风速大于切出风速
            Pwt_samp(i)=0;%风电功率等于0
        end
    end