Motion forecasting for on-road traffic agents presents both a significant challenge and a critical necessity for ensuring safety in autonomous driving systems. In contrast to most existing data-driven approaches that directly predict future trajectories, we rethink this task from a planning perspective, advocating a "First Reasoning, Then Forecasting" strategy that explicitly incorporates behavior intentions as spatial guidance for trajectory prediction. To achieve this, we introduce an interpretable, reward-driven intention reasoner grounded in a novel query-centric Inverse Reinforcement Learning (IRL) scheme. Our method first encodes traffic agents and scene elements into a unified vectorized representation, then aggregates contextual features through a query-centric paradigm. This enables the derivation of a reward distribution, a compact yet informative representation of the target agent's behavior within the given scene context via IRL. Guided by this reward heuristic, we perform policy rollouts to reason about multiple plausible intentions, providing valuable priors for subsequent trajectory generation. Finally, we develop a hierarchical DETR-like decoder integrated with bidirectional selective state space models to produce accurate future trajectories along with their associated probabilities. Extensive experiments on the large-scale Argoverse and nuScenes motion forecasting datasets demonstrate that our approach significantly enhances trajectory prediction confidence, achieving highly competitive performance relative to state-of-the-art methods.

Trajectory prediction for surrounding agents is a challenging task in autonomous driving due to its inherent uncertainty and underlying multimodality. Unlike prevailing data-driven methods that primarily rely on supervised learning, in this paper, we introduce a novel Graph-oriented Inverse Reinforcement Learning (GoIRL) framework, which is an IRL-based predictor equipped with vectorized context representations. We develop a feature adaptor to effectively aggregate lane-graph features into grid space, enabling seamless integration with the maximum entropy IRL paradigm to infer the reward distribution and obtain the policy that can be sampled to induce multiple plausible plans. Furthermore, conditioned on the sampled plans, we implement a hierarchical parameterized trajectory generator with a refinement module to enhance prediction accuracy and a probability fusion strategy to boost prediction confidence. Extensive experimental results showcase our approach not only achieves state-of-the-art performance on the large-scale Argoverse & nuScenes motion forecasting benchmarks but also exhibits superior generalization abilities compared to existing supervised models.

This article deals with the problem of mobile robot path planning in an unknown environment that contains both static and dynamic obstacles, utilizing a reinforcement learning approach. We propose an improved Dyna-${Q}$algorithm, which incorporates heuristic search strategies, simulated annealing mechanism, and reactive navigation principle into${Q}$-learning based on the Dyna architecture. A novel action-selection strategy combining$\varepsilon $-greedy policy with the cooling schedule control is presented, which, together with the heuristic reward function and heuristic actions, can tackle the exploration-exploitation dilemma and enhance the performance of global searching, convergence property, and learning efficiency for path planning. The proposed method is superior to the classical${Q}$-learning and Dyna-${Q}$algorithms in an unknown static environment, and it is successfully applied to an uncertain environment with multiple dynamic obstacles in simulations. Further, practical experiments are conducted by integrating MATLAB and robot operating system (ROS) on a physical robot platform, and the mobile robot manages to find a collision-free path, thus fulfilling autonomous navigation tasks in the real world.