L2RPNSandBoxScore.py

# Copyright (c) 2019-2020, RTE (https://www.rte-france.com)
# See AUTHORS.txt
# This Source Code Form is subject to the terms of the Mozilla Public License, version 2.0.
# If a copy of the Mozilla Public License, version 2.0 was not distributed with this file,
# you can obtain one at http://mozilla.org/MPL/2.0/.
# SPDX-License-Identifier: MPL-2.0
# This file is part of Grid2Op, Grid2Op a testbed platform to model sequential decision making in power systems.

import numpy as np

from grid2op.Reward.BaseReward import BaseReward
from grid2op.dtypes import dt_float


class L2RPNSandBoxScore(BaseReward):
    """
    INTERNAL

    .. warning:: /!\\\\ Internal, do not use unless you know what you are doing /!\\\\
            It **must not** serve as a reward. This scored needs to be minimized,
            and a reward needs to be maximized! Also, this "reward" is not scaled or anything. Use it as your
            own risk.

    Implemented as a reward to make it easier to use in the context of the L2RPN competitions, this "reward"
    computed the "grid operation cost". It should not be used to train an agent.

    The "reward" the closest to this score is given by the :class:`RedispReward` class.

    """
    def __init__(self, alpha_redisph=1.0):
        BaseReward.__init__(self)
        self.reward_min = dt_float(1.0)  # carefull here between min and max...
        self.reward_max = dt_float(300.0 * 70.0)
        self.alpha_redisph = dt_float(alpha_redisph)

    def __call__(self,  action, env, has_error, is_done, is_illegal, is_ambiguous):
        if has_error:
            # DO SOMETHING IN THIS CASE
            return self.reward_min

        # compute the losses
        gen_p, *_ = env.backend.generators_info()
        load_p, *_ = env.backend.loads_info()
        losses = np.sum(gen_p, dtype=dt_float) - np.sum(load_p, dtype=dt_float)

        # compute the marginal cost
        gen_activeprod_t = env._gen_activeprod_t
        p_t = np.max(env.gen_cost_per_MW[gen_activeprod_t > 0.]).astype(dt_float)

        # redispatching amount
        actual_dispatch = env._actual_dispatch
        c_redispatching = dt_float(2.0) * self.alpha_redisph * np.sum(np.abs(actual_dispatch)) * p_t

        # cost of losses
        c_loss = losses * p_t

        # total "operationnal cost"
        c_operations = dt_float(c_loss + c_redispatching)

        return c_operations