update heat and grid part

models/env.py

@@ -20,7 +20,7 @@ class WgzGym(gym.Env):
         self.TRAIN = True
         self.current_time = None
         self.episode_length = 24
-        self.penalty_coefficient = 50  # 约束惩罚系数
+        self.penalty_coefficient = 10  # 约束惩罚系数
         self.sell_coefficient = 0.5  # 售出利润系数
         self.a = 0.5
         self.b = 0.3
@@ -67,7 +67,8 @@ class WgzGym(gym.Env):
 
         obs = np.concatenate((np.float32(time_step), np.float32(price), np.float32(temper),
                               np.float32(solar), np.float32(load), np.float32(heat),
-                              np.float32(people), np.float32(ec_out), np.float32(hst_soc), np.float32(grid_ex)), axis=None)
+                              np.float32(people), np.float32(ec_out), np.float32(hst_soc), np.float32(grid_ex)),
+                             axis=None)
         return obs
 
     def step(self, action):
@@ -82,17 +83,26 @@ class WgzGym(gym.Env):
         load = current_obs[4]
         heat = current_obs[5]
 
-        gym_to_grid = solar + self.HST.get_power() - self.EC.current_power - load
-        heat_penalty = 0
+        power_gap = solar + self.HST.get_power() - self.EC.current_power - load
+        heat_gap = self.HST.get_heat() + self.EC.get_heat() - heat
 
         # reward = 0.0
         sell_benefit, buy_cost = 0, 0
-        # power_penalty = 0
-        if gym_to_grid >= 0:  # 过剩
-            sell_benefit = self.grid.get_cost(price, gym_to_grid) * self.sell_coefficient
+        if power_gap >= 0:  # 过剩
+            sell_benefit = self.grid.get_cost(price, power_gap) * self.sell_coefficient
+            power_gap = 0
+            power_penalty = 0
         else:  # 缺少
-            buy_cost = self.grid.get_cost(price, abs(gym_to_grid))
-        power_penalty = abs(gym_to_grid) * self.penalty_coefficient
+            power_gap = abs(power_gap)
+            buy_cost = self.grid.get_cost(price, power_gap)
+            power_penalty = power_gap * self.penalty_coefficient
+
+        if heat_gap >= 0:
+            heat_gap = 0
+            heat_penalty = 0
+        else:
+            heat_gap = abs(heat_gap)
+            heat_penalty = heat_gap * self.penalty_coefficient
 
         hst_cost = self.HST.get_cost()
         ec_cost = self.EC.get_cost(price)
@@ -100,10 +110,10 @@ class WgzGym(gym.Env):
 
         economic_cost = hst_cost + ec_cost + solar_cost - sell_benefit + buy_cost
         demand_cost = self.heat_a * heat_penalty + self.power_a * power_penalty
-        eco_benifit = 0
+        eco_benifit = self.EC.less_carbon() - self.grid.get_carbon()
         reward = - self.a * demand_cost - self.b * economic_cost + self.c * eco_benifit
 
-        self.unbalance = gym_to_grid
+        self.unbalance = power_gap + heat_gap
         final_step_outputs = [self.HST.current_soc, self.EC.current_power, self.grid.current_power]
         self.current_time += 1
         finish = (self.current_time == self.episode_length)

models/module.py

@@ -25,7 +25,7 @@ class EC:
     def get_heat(self):
         return self.current_power * (1 - self.electricity_efficiency)
 
-    def get_carbon(self):
+    def less_carbon(self):
         return self.current_power * self.carbon_reduce
 
     def get_power_max(self):
@@ -84,6 +84,7 @@ class HST:
 class Grid:
     def __init__(self):
         self.delta = 1
+        self.carbon_increace = 0.9
         self.trade_energy = None
 
     def step(self, action_grid):
@@ -92,6 +93,9 @@ class Grid:
     def get_cost(self, price):
         return price * self.trade_energy * self.delta
 
+    def get_carbon(self):
+        return self.trade_energy * self.carbon_increace
+
     def retrieve_past_price(self):
         result = []
         # 过去24小时的价格起始、结束索引

models/parameters.py

@@ -6,7 +6,7 @@ EC_parameters = {
     'lifetime': 6000,  # hour
     'equipment_cost': 10000,  # yuan
     'electrolysis_efficiency': 0.8,
-    'carbon_reduce': 1,
+    'carbon_reduce': 0.9,
 }
 
 HST_parameters = {