Optimal Decision-Making Method for a Plug-In Electric Taxi in Uncertain Environment

Abstract

This paper studies the optimal decision-making problem for a plug-in electric taxi (PET) in a time-varying complex environment, i.e., a passenger environment, charging station environment, traffic environment, and taxi company management system, in order to maximize PET profit in a short-term operating cycle. First, this problem is formulated as a sequential decision-making problem composed of multiple decision slots. Then, to make the model more practical, the model is divided into two parts: an external environment and an electric taxi model for refinement. The uncertainty and time-varying characteristics of four environmental aspects, including passengers, charging stations, traffic, and taxi company management systems, are analysed and modelled. The transitions between adjacent processes and the environmental feedback of each process are modelled by further subdividing both the serving process and the charging process of the PET into multiple subprocesses, including cruising, carrying passengers, driving to the charging station, queueing before charging, and connecting to the power grid for charging. There are several uncertain factors in the sequential decision-making process for the PET, which leads to difficulty in solving the problem. To address this difficulty, the model-free algorithm SARSA is chosen. Finally, the effectiveness of the proposed method is verified by simulation results.INDEX TERMS Plug-in electric taxi, decision making, uncertainty, SARSA algorithm, load modelling.

Empirical analysis of decision recommendation models for various processes from a pragmatic perspective

“…First, this is a lengthy sequential decision-making issue. For convenience, the model is separated into an outdoor scene and an electric taxi model You et al (2021). The management methods used by taxi companies, charging stations, traffic, and passengers are among the four environmental aspects that have been analyzed and modeled.…”

Planning of distributed renewable energy systems under uncertainty based on statistical machine learning

“…The uncertainty of passengers, charging stations, and public transport usually leads to the uncertainty of electric taxi loads. Therefore, the operational behavior model of electric plug-in taxis (PET) is presented based on environmental uncertainty [85]. In [86], regarding the movements of electric taxis as random walks, the Markov process is used to simulate the distribution of charging demand in static space, while in [87], the random forest (RF) method based on a regression tree is used to predict the driving characteristics of each EV, so as to obtain the travel mode data set of the EV.…”

Smart Carbon Emission Scheduling for Electric Vehicles via Reinforcement Learning under Carbon Peak Target

“…Reinforcement learning aims to find an optimal policy to make the agent obtain the maximum reward by taking the action according to the current state. Sarsa [25] and Q-learning [26] estimate the future reward of a policy by utilizing temporal difference iteration and then choose the discrete action drawn from the highest reward. However, it is difficult to set the convincing discrete action.…”

“…, a(t + 1)) 18: end 19: Update ε using dε and of ω using dω according to (24) and (25). 20: Until k > ξ max 21: End RLSCA has the policy Θ(a t |ϕ t ; ε) and the approximate state function Q(ϕ t ; ω) and adopts the multi-step returns [19] to update the strategy and the value function after every t max actions.…”