Source code for src.datafev.algorithms.vehicle.routing_milp

# The datafev framework

# Copyright (C) 2022,
# Institute for Automation of Complex Power Systems (ACS),
# E.ON Energy Research Center (E.ON ERC),
# RWTH Aachen University

# Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated
# documentation files (the "Software"), to deal in the Software without restriction, including without limitation the
# rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit
# persons to whom the Software is furnished to do so, subject to the following conditions:

# The above copyright notice and this permission notice shall be included in all copies or substantial portions of the
# Software.

# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
# WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
# COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
# OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.


from pyomo.core import *
import pyomo.kernel as pmo


[docs]def smart_routing( solver, opt_horizon, opt_step, ecap, v2gall, tarsoc, minsoc, maxsoc, crtsoc, crttime, arrtime, deptime, arrsoc, p_ch, p_ds, g2v_dps, v2g_dps, ): """ This function optimizes: - the allocation of an incoming EV to a cluster, - and the charging schedule in the given parking duration considering cluster differentiated dynamic price signals. Parameters ---------- opt_step : float Size of one time step in the optimization (seconds). opt_horizon : list of integers Time step identifiers in the optimization horizon. ecap : float Energy capacity of battery (kWs). v2gall : float V2G allowance discharge (kWs). tarsoc : float Target final soc (0<inisoc<1). minsoc : float Minimum soc. maxsoc : float Maximum soc. crtsoc : float Target soc at crttime. crttime : int Critical time s.t. s(srttime)> crtsoc. arrtime : dict of int Cluster differentiating arrival times. deptime : dict of int Cluster differentiating departure times. arrsoc : dict of float Cluster differentiating arrival soc \in [0,1). p_ch : dict of float Nominal charging power (kW). p_ds : dict of float Nominal charging power (kW). g2v_dps : dict of dict G2V dynamic price signals of clusters (Eur/kWh). v2g_dps : dict of dict V2G dynamic price signals of clusters (Eur/kWh). Returns ------- p_schedule : dict Power schedule. Each item in the EV dictionary indicates the power to be supplied to the EV(kW) during a particular time step. s_schedule : dict SOC schedule. Each item in the EV dictionary indicates the SOC to be achieved by the EV by a particular time step. target_cc : string Cluster to send the EV. """ conf_period = {} for t in opt_horizon: if t < crttime: conf_period[t] = 0 else: conf_period[t] = 1 candidate_clusters = arrtime.keys() ####################Constructing the optimization model#################### model = ConcreteModel() model.T = Set(initialize=opt_horizon, ordered=True) # Time index set model.C = Set(initialize=candidate_clusters, ordered=True) # Cluster index set model.dt = opt_step # Step size model.E = ecap # Battery capacity in kWs model.SoC_F = tarsoc # SoC to be achieved at the end model.SoC_R = crtsoc # Minimim SOC must be ensured in the confidence period model.conf = conf_period # Confidence period where SOC must be larger than crtsoc model.V2G_ALL = v2gall # Maximum energy that can be discharged V2G model.P_CH_Max = max(p_ch.values()) # Maximum available charging power in kW model.P_DS_Max = max(p_ds.values()) # Maximum available discharging power in kW model.P_CH = p_ch # Cluster dependent max charging power in kW model.P_DS = p_ds # Cluster dependent max discharging power in kW model.W_G2V = g2v_dps # Time-variant G2V cost coefficients of clusters model.W_V2G = v2g_dps # Time-variant V2G cost coefficients of clusters model.t_arr = arrtime # Cluster dependent arrival time estimation model.t_dep = deptime # Cluster dependent departure time estimation model.SoC_I = arrsoc # Cluster dependent arrival SOCs estimation model.xc = Var( model.C, within=pmo.Binary ) # Binary variable having 1 if v is allocated to c model.xp = Var( model.T, within=pmo.Binary ) # Binary variable having 1/0 if v is charged/discharged at t model.p = Var(model.T, within=Reals) # Net charge power at t model.p_pos = Var(model.T, within=NonNegativeReals) # Charge power at t model.p_neg = Var(model.T, within=NonNegativeReals) # Discharge power at t model.pc_pos = Var( model.C, model.T, within=NonNegativeReals ) # Charge power at t if it is in cluster c model.pc_neg = Var( model.C, model.T, within=NonNegativeReals ) # Discharge power at t if it is in cluster c model.SoC = Var( model.T, within=NonNegativeReals, bounds=(minsoc, maxsoc) ) # SOC to be achieved at time step t # CONSTRAINTS def initialsoc(model): return model.SoC[0] == sum(model.xc[c] * model.SoC_I[c] for c in model.C) model.inisoc = Constraint(rule=initialsoc) def storageConservation( model, t ): # SOC of EV batteries will change with respect to the charged power and battery energy capacity if t < max(model.T): return model.SoC[t + 1] == (model.SoC[t] + model.p[t] * model.dt / model.E) else: return model.SoC[t] == model.SoC_F model.socconst = Constraint(model.T, rule=storageConservation) def socconfidence(model, t): return model.SoC[t] >= model.SoC_R * model.conf[t] model.socconfi = Constraint(model.T, rule=socconfidence) def supplyrule_end(model): return model.p[max(model.T)] == 0.0 model.supconst = Constraint(rule=supplyrule_end) def combinatorics0(model): # EV can assigned to only one cluster return sum(model.xc[c] for c in model.C) == 1 model.comb0const = Constraint(rule=combinatorics0) def combinatorics11(model, c, t): if model.t_arr[c] <= t < model.t_dep[c]: return model.pc_neg[c, t] <= model.P_DS[c] * model.xc[c] else: return model.pc_neg[c, t] == 0 model.comb11const = Constraint(model.C, model.T, rule=combinatorics11) def combinatorics12(model, c, t): if model.t_arr[c] <= t < model.t_dep[c]: return model.pc_pos[c, t] <= model.P_CH[c] * model.xc[c] else: return model.pc_pos[c, t] == 0 model.comb12const = Constraint(model.C, model.T, rule=combinatorics12) def combinatorics2(model, t): return model.p[t] == sum( model.pc_pos[c, t] - model.pc_neg[c, t] for c in model.C ) model.comb2const = Constraint(model.T, rule=combinatorics2) def netcharging(model, t): return model.p[t] == model.p_pos[t] - model.p_neg[t] model.netchr = Constraint(model.T, rule=netcharging) def combinatorics31_pos(model, t): return model.p_pos[t] <= model.xp[t] * model.P_CH_Max model.comb31pconst = Constraint(model.T, rule=combinatorics31_pos) def combinatorics32_pos(model, t): return model.p_pos[t] == sum(model.pc_pos[c, t] for c in model.C) model.comb32pconst = Constraint(model.T, rule=combinatorics32_pos) def combinatorics31_neg(model, t): return model.p_neg[t] <= (1 - model.xp[t]) * model.P_DS_Max model.comb31nconst = Constraint(model.T, rule=combinatorics31_neg) def combinatorics32_neg(model, t): return model.p_neg[t] == sum(model.pc_neg[c, t] for c in model.C) model.comb32nconst = Constraint(model.T, rule=combinatorics32_neg) def v2g_limit(model): return sum(model.p_neg[t] * model.dt for t in model.T) <= model.V2G_ALL model.v2gconst = Constraint(rule=v2g_limit) # OBJECTIVE FUNCTION def obj_rule(model): return ( sum( model.W_G2V[c][t] * model.pc_pos[c, t] - model.W_V2G[c][t] * model.pc_neg[c, t] for c in model.C for t in opt_horizon[:-1] ) * opt_step / 3600 ) model.obj = Objective(rule=obj_rule, sense=minimize) # model.pprint() result = solver.solve(model) # ,tee=True) # print(result) p_schedule = {} s_schedule = {} for t in model.T: p_schedule[t] = model.p[t]() s_schedule[t] = model.SoC[t]() for c in model.C: if abs(model.xc[c]() - 1) <= 0.01: target_cc = c return p_schedule, s_schedule, target_cc
if __name__ == "__main__": import pandas as pd import numpy as np from pyomo.environ import * from pyomo.opt import SolverFactory ########################################################################### # Input parameters solver = SolverFactory("cplex") opt_step = 300 # seconds opt_horizon = range(13) # [0 1 2 3 4 .. 12] == 1 hour for opt_step=300 seconds ecap = 50 * 3600 # kWs v2gall = 10 * 3600 # kWs tarsoc = 1.0 minsoc = 0.4 maxsoc = 1.0 crtsoc = tarsoc crttime = 12 arrtime = {"C1": 0, "C2": 5, "C3": 5} deptime = {"C1": 13, "C2": 13, "C3": 13} arrsoc = {"C1": 0.5, "C2": 0.49, "C3": 0.49} p_ch = {"C1": 50, "C2": 50, "C3": 50} p_ds = {"C1": 50, "C2": 50, "C3": 50} np.random.seed(0) g2v_dps = {} v2g_dps = {} for c in ["C1", "C2", "C3"]: g2v_tariff = np.random.uniform(low=0.4, high=0.8, size=12) g2v_dps[c] = dict(enumerate(g2v_tariff)) v2g_dps[c] = dict(enumerate(g2v_tariff * 0.9)) dps = {} for c in ["C1", "C2", "C3"]: dps[c] = pd.DataFrame(columns=["G2V", "V2G"]) dps[c]["G2V"] = pd.Series(g2v_dps[c]) dps[c]["V2G"] = pd.Series(v2g_dps[c]) ########################################################################### print("Reservation request of an EV with") print("Battery capacity :", ecap / 3600, "kWh") print("Target SOC :", tarsoc) print("V2G allowance :", v2gall / 3600, "kWh") print() print( "Since the available clusters are at different distances, some parameters are cluster dependent" ) print("Estimated arrival SOCs :", arrsoc) print("Estimated arrival time steps :", arrtime) print("Estimated departure time steps:", deptime) print("Dynamic price signals of the clusters:") print(pd.concat(dps, axis=1)) print() print("Smart routing optimization problem is solved...") p, s, c = smart_routing( solver, opt_horizon, opt_step, ecap, v2gall, tarsoc, minsoc, maxsoc, crtsoc, crttime, arrtime, deptime, arrsoc, p_ch, p_ds, g2v_dps, v2g_dps, ) print() print("The result:") print( "Under the given price signals, the optimal decision is to go to the cluster", c ) print() print("And charge with the profile is printed in table") print("SOC (%): SOC trajectory in optimized schedule") print("P (kW): Power supply to the EV in optimized schedule") results = pd.DataFrame(columns=["P (kW)", "SOC (%)"], index=sorted(s.keys())) results["P (kW)"] = pd.Series(p) results["SOC (%)"] = pd.Series(s) * 100 print(results) print()