arXiv subjects

Guojian Zhan

Publications and source records attributed to Guojian Zhan.

25 records · Page 2Linked to original sources

arXiv2025

Predictive Lagrangian Optimization for Constrained Reinforcement Learning

Constrained optimization is popularly seen in reinforcement learning for addressing complex control tasks. From the perspective of dynamic system, iteratively solving a constrained optimization problem can be framed as the temporal evolution of a feedback control system. Classical constrained optimization methods, such as penalty and Lagrangian approaches, inherently use proportional and integral feedback controllers. In this paper, we propose a more generic equivalence framework to build the connection between constrained optimization and feedback control system, for the purpose of developing more effective constrained RL algorithms. Firstly, we define that each step of the system evolution determines the Lagrange multiplier by solving a multiplier feedback optimal control problem (MFOCP). In this problem, the control input is multiplier, the state is policy parameters, the dynamics is described by policy gradient descent, and the objective is to minimize constraint violations. Then, we introduce a multiplier guided policy learning (MGPL) module to perform policy parameters updating. And we prove that the resulting optimal policy, achieved through alternating MFOCP and MGPL, aligns with the solution of the primal constrained RL problem, thereby establishing our equivalence framework. Furthermore, we point out that the existing PID Lagrangian is merely one special case within our framework that utilizes a PID controller. We also accommodate the integration of other various feedback controllers, thereby facilitating the development of new algorithms. As a representative, we employ model predictive control (MPC) as the feedback controller and consequently propose a new algorithm called predictive Lagrangian optimization (PLO). Numerical experiments demonstrate its superiority over the PID Lagrangian method, achieving a larger feasible region up to 7.2% and a comparable average reward.

cs.LG↗

arXiv2025

Zeroth-Order Actor-Critic: An Evolutionary Framework for Sequential Decision Problems

Evolutionary algorithms (EAs) have shown promise in solving sequential decision problems (SDPs) by simplifying them to static optimization problems and searching for the optimal policy parameters in a zeroth-order way. While these methods are highly versatile, they often suffer from high sample complexity due to their ignorance of the underlying temporal structures. In contrast, reinforcement learning (RL) methods typically formulate SDPs as Markov Decision Process (MDP). Although more sample efficient than EAs, RL methods are restricted to differentiable policies and prone to getting stuck in local optima. To address these issues, we propose a novel evolutionary framework Zeroth-Order Actor-Critic (ZOAC). We propose to use step-wise exploration in parameter space and theoretically derive the zeroth-order policy gradient. We further utilize the actor-critic architecture to effectively leverage the Markov property of SDPs and reduce the variance of gradient estimators. In each iteration, ZOAC employs samplers to collect trajectories with parameter space exploration, and alternates between first-order policy evaluation (PEV) and zeroth-order policy improvement (PIM). To evaluate the effectiveness of ZOAC, we apply it to a challenging multi-lane driving task, optimizing the parameters in a rule-based, non-differentiable driving policy that consists of three sub-modules: behavior selection, path planning, and trajectory tracking. We also compare it with gradient-based RL methods on three Gymnasium tasks, optimizing neural network policies with thousands of parameters. Experimental results demonstrate the strong capability of ZOAC in solving SDPs. ZOAC significantly outperforms EAs that treat the problem as static optimization and matches the performance of gradient-based RL methods even without first-order information, in terms of total average return across all tasks.

cs.LG↗

arXiv2024

An Explicit Discrete-Time Dynamic Vehicle Model with Assured Numerical Stability

Numerical stability is of great significance for discrete-time dynamic vehicle model. Among the unstable factors, low-speed singularity stands out as one of the most challenging issues, which arises from that the denominator of tire side angle term only contains the vehicle longitudinal speed. Consequently, for the common low-speed and stop-start driving scenarios, the calculated tire slip angle will approach infinity, which will further lead to the numerical explosion of other vehicle states. In response to this critical challenge, we propose a discrete-time dynamic vehicle model that effectively mitigates the low-speed singularity issue, ensuring numerical stability and maintaining the explicit form-highly favored by model-based control algorithms. To validate the numerical stability of our model, we conduct a rigorous theoretical analysis, establishing sufficient conditions for stability, and conduct extensive empirical validation tests across a wide spectrum of speeds. Subsequent to the validation process, we conduct comprehensive simulations comparing our proposed model with both kinematic models and existing dynamic models discretized through the forward Euler method. The results demonstrate that our proposed model shows better comprehensive performance in terms of both the accuracy and numerical stability. Finally, the real vehicle experiments are carried out to support that our proposed model can closely aligns to the real vehicle trajectories showcasing its practicality and ease of use. Notably, our work stands as the pioneering endeavor in introducing an explicit discrete-time dynamic vehicle model suitable for common urban driving scenarios including low-speed and stop-start.

eess.SY↗

arXiv2024

Rocket Landing Control with Random Annealing Jump Start Reinforcement Learning

Rocket recycling is a crucial pursuit in aerospace technology, aimed at reducing costs and environmental impact in space exploration. The primary focus centers on rocket landing control, involving the guidance of a nonlinear underactuated rocket with limited fuel in real-time. This challenging task prompts the application of reinforcement learning (RL), yet goal-oriented nature of the problem poses difficulties for standard RL algorithms due to the absence of intermediate reward signals. This paper, for the first time, significantly elevates the success rate of rocket landing control from 8% with a baseline controller to 97% on a high-fidelity rocket model using RL. Our approach, called Random Annealing Jump Start (RAJS), is tailored for real-world goal-oriented problems by leveraging prior feedback controllers as guide policy to facilitate environmental exploration and policy learning in RL. In each episode, the guide policy navigates the environment for the guide horizon, followed by the exploration policy taking charge to complete remaining steps. This jump-start strategy prunes exploration space, rendering the problem more tractable to RL algorithms. The guide horizon is sampled from a uniform distribution, with its upper bound annealing to zero based on performance metrics, mitigating distribution shift and mismatch issues in existing methods. Additional enhancements, including cascading jump start, refined reward and terminal condition, and action smoothness regulation, further improve policy performance and practical applicability. The proposed method is validated through extensive evaluation and Hardware-in-the-Loop testing, affirming the effectiveness, real-time feasibility, and smoothness of the proposed controller.

cs.LG↗

arXiv2024

Canonical Form of Datatic Description in Control Systems

The design of feedback controllers is undergoing a paradigm shift from modelic (i.e., model-driven) control to datatic (i.e., data-driven) control. Canonical form of state space model is an important concept in modelic control systems, exemplified by Jordan form, controllable form and observable form, whose purpose is to facilitate system analysis and controller synthesis. In the realm of datatic control, there is a notable absence in the standardization of data-based system representation. This paper for the first time introduces the concept of canonical data form for the purpose of achieving more effective design of datatic controllers. In a control system, the data sample in canonical form consists of a transition component and an attribute component. The former encapsulates the plant dynamics at the sampling time independently, which is a tuple containing three elements: a state, an action and their corresponding next state. The latter describes one or some artificial characteristics of the current sample, whose calculation must be performed in an online manner. The attribute of each sample must adhere to two requirements: (1) causality, ensuring independence from any future samples; and (2) locality, allowing dependence on historical samples but constrained to a finite neighboring set. The purpose of adding attribute is to offer some kinds of benefits for controller design in terms of effectiveness and efficiency. To provide a more close-up illustration, we present two canonical data forms: temporal form and spatial form, and demonstrate their advantages in reducing instability and enhancing training efficiency in two datatic control systems.

eess.SY↗

arXiv2022

Quantifying the Individual Differences of Driver' Risk Perception with Just Four Interpretable Parameters

There will be a long time when automated vehicles are mixed with human-driven vehicles. Understanding how drivers assess driving risks and modelling their individual differences are significant for automated vehicles to develop human-like and customized behaviors, so as to gain people's trust and acceptance. However, the reality is that existing driving risk models are developed at a statistical level, and no one scenario-universal driving risk measure can correctly describe risk perception differences among drivers. We proposed a concise yet effective model, called Potential Damage Risk (PODAR) model, which provides a universal and physically meaningful structure for driving risk estimation and is suitable for general non-collision and collision scenes. In this paper, based on an open-accessed dataset collected from an obstacle avoidance experiment, four physical-interpretable parameters in PODAR, including prediction horizon, damage scale, temporal attenuation, and spatial attention, are calibrated and consequently individual risk perception models are established for each driver. The results prove the capacity and potential of PODAR to model individual differences in perceived driving risk, laying the foundation for autonomous driving to develop human-like behaviors.

math.NA↗

arXiv2022

Improve Generalization of Driving Policy at Signalized Intersections with Adversarial Learning

Intersections are quite challenging among various driving scenes wherein the interaction of signal lights and distinct traffic actors poses great difficulty to learn a wise and robust driving policy. Current research rarely considers the diversity of intersections and stochastic behaviors of traffic participants. For practical applications, the randomness usually leads to some devastating events, which should be the focus of autonomous driving. This paper introduces an adversarial learning paradigm to boost the intelligence and robustness of driving policy for signalized intersections with dense traffic flow. Firstly, we design a static path planner which is capable of generating trackable candidate paths for multiple intersections with diversified topology. Next, a constrained optimal control problem (COCP) is built based on these candidate paths wherein the bounded uncertainty of dynamic models is considered to capture the randomness of driving environment. We propose adversarial policy gradient (APG) to solve the COCP wherein the adversarial policy is introduced to provide disturbances by seeking the most severe uncertainty while the driving policy learns to handle this situation by competition. Finally, a comprehensive system is established to conduct training and testing wherein the perception module is introduced and the human experience is incorporated to solve the yellow light dilemma. Experiments indicate that the trained policy can handle the signal lights flexibly meanwhile realizing the smooth and efficient passing with a humanoid paradigm. Besides, APG enables a large-margin improvement of the resistance to the abnormal behaviors and thus ensures a high safety level for the autonomous vehicle.

cs.RO↗