基于Python Pandas的电池SOC简单模型

Question

我在使用Pandas进行数据分析方面相对较新。

我试图用几种条件定义一个函数，然后使用apply()函数来建模电池存储系统的充电状态(SOC)。

由于我需要访问以前的SOC，所以在我的功能中使用shift(1)。

我真的试图使它简单，但我仍然不知道如何限制电池SOC为16 kWh，并使我的代码运行。

我有以下Pandas Dataframe:带有太阳能电池板的房屋的电力概况

然后，我在Pandas Dataframe中创建了另外两个列“盈余”和"SOC“。

1表示电力生产盈余，0表示电力负荷较高。

Electricity_profiles_df['surplus'] = Electricity_profiles_df['Electricity Surplus / Deficit [kWh]'].apply(lambda x: 1 if x > 0 else 0)

Electricity_profiles_df['SOC'] = np.NaN

然后，我将我的功能定义如下：

def state_of_charge(surplus, SOC, power):
if surplus == 1:
    if SOC.shift(1) < 16:
        SOC = power + SOC.shift(1)
    else:
        SOC = SOC.shift(1)
elif surplus == 0:
    if SOC.shift(1) > 3.2:
        SOC = SOC.shift(1) + power
    else:
        SOC = SOC.shift(1)
else:
    SOC = 3.2
return SOC

最后，我尝试使用新定义的函数，并按行顺序使用apply()函数：

Electricity_profiles_df['SOC'] = Electricity_profiles_df['surplus','SOC', 'Electricity Surplus / Deficit [kWh]'].apply(state_of_charge, axis=1)

显示以下错误：

KeyError

非常感谢你的帮助！

Answer 1

在木星笔记本上试试这个。基本上，您在numpy数组中工作，并且只在最后将这些数组添加到一个熊猫数据数组中。

import pandas as pd
import numpy as np

pd.options.plotting.backend = "plotly"

%%time
#Parameters
SOC0 = 0
sim_time = 90
A_demand = 1
A_prod = 1.2
max_SOC = 16

#Signals generation

    #Initial empty arrays
t = np.arange(0,sim_time)
grid_demand =  np.zeros(sim_time)
battery = np.zeros(sim_time)
charge_array = np.zeros(sim_time)

    #External production and demand signals (can set any shape of curve)
prod = (np.sin(t/(2*np.pi))*A_prod + A_prod)/2
demand = np.ones(sim_time)*A_demand

#Estimates battery SOC
SOC = SOC0
for i in t:

   battery[i] = SOC

    #Energy required from the grid (No surplus)
    if demand[i] - prod[i] >= 0:
    
        #Battery has some charge left for next step
        if (demand[i] - prod[i] - SOC) < 0 :
            grid_demand[i] = 0
            charge = -(demand[i] - prod[i])
    
        #Battery fully discharges
        else:
            grid_demand[i] = demand[i] - prod[i] - SOC
            charge = -SOC

    #Surplus of energy demand
    elif demand[i] - prod[i] < 0:
    
        grid_demand[i] = 0
        charge = prod[i] - demand[i]

    SOC = min(SOC + charge,max_SOC)
    charge_array[i] = charge
    
#Creates data frame
data_dict = {"demand":demand,
            "prod":prod,
            "battery":battery,
            "grid_demand":grid_demand,
            "charge":charge_array}

df = pd.DataFrame(data_dict)

#Plots
df.plot()