Mohammad Zamani

Mohammad Zamani, Valery Ugrinovskii

The paper addresses the problem of distributed filtering with guaranteed convergence properties using minimum-energy filtering and $H_\infty$ filtering methodologies. A linear state space plant model is considered observed by a network of communicating sensors, in which individual sensor measurements may lead to an unobservable filtering problem. However, each filter locally shares estimates, that are subject to disturbances, with its respective neighboring filters to produce an estimate of the plant state. The minimum-energy strategy of the proposed local filter leads to a locally optimal time-varying filter gain facilitating the transient and the asymptotic convergence of the estimation error, with guaranteed $H_\infty$ performance. The filters are implementable using only the local measurements and information from the neighboring filters subject to disturbances. A key idea of the proposed algorithm is to locally approximate the neighboring estimates, that are not directly accessible, considering them as disturbance contaminated versions of the plant state. The proposed algorithm imposes minimal communication load on the network and is scalable to larger sensor networks.

Mohammad Zamani, Iman Shames, Valery Ugrinovskii

This paper presents a consensus algorithm for a multi-agent system where each agent has access to its imperfect own state and neighboring state measurements. The measurements are subject to deterministic disturbances and the proposed algorithm provides a minimum-energy estimate of the measured states which is instrumental in achieving consensus by the nodes. It is shown that the proposed consensus algorithm converges exponentially in the absence of disturbances, and its performance under bounded continuous disturbances is investigated as well. The convergence performance of the proposed method is further studied using simulations where we show that consensus is achieved despite using large measurement errors.

M. Zamani, V. Ugrinovskii

The paper considers the problem of cooperative estimation for a linear uncertain plant observed by a network of communicating sensors. We take a novel approach by treating the filtering problem from the view point of local sensors while the network interconnections are accounted for via an uncertain signals modelling of estimation performance of other nodes. That is, the information communicated between the nodes is treated as the true plant information subject to perturbations, and each node is endowed with certain believes about these perturbations during the filter design. The proposed distributed filter achieves a suboptimal $H_\infty$ consensus performance. Furthermore, local performance of each estimator is also assessed given additional constraints on the performance of the other nodes. These conditions are shown to be useful in tuning the desired estimation performance of the sensor network.

Oliver Biggar, Mohammad Zamani, Iman Shames

In this paper we provide a formal framework for comparing the expressive power of Behavior Trees (BTs) to other action selection architectures. Taking inspiration from the analogous comparisons of structural programming methodologies, we formalise the concept of `expressiveness'. This leads us to an expressiveness hierarchy of control architectures, which includes BTs, Decision Trees (DTs), Teleo-reactive Programs (TRs) and Finite State Machines (FSMs). By distinguishing between BTs with auxiliary variables and those without, we demonstrate the existence of a trade-off in BT design between readability and expressiveness. We discuss what this means for BTs in practice.

Jack Henderson, Mohammad Zamani, Robert Mahony et al.

Accurate localisation of unmanned aerial vehicles is vital for the next generation of automation tasks. This paper proposes a minimum energy filter for velocity-aided pose estimation on the extended special Euclidean group. The approach taken exploits the Lie-group symmetry of the problem to combine Inertial Measurement Unit (IMU) sensor output with landmark measurements into a robust and high performance state estimate. We propose an asynchronous discrete-time implementation to fuse high bandwidth IMU with low bandwidth discrete-time landmark measurements typical of real-world scenarios. The filter's performance is demonstrated by simulation.

Oliver Biggar, Mohammad Zamani, Iman Shames

As complex autonomous robotic systems become more widespread, the need for transparent and reusable Artificial Intelligence (AI) designs becomes more apparent. In this paper we analyse how the principles behind Behavior Trees (BTs), an increasingly popular tree-structured control architecture, are applicable to these goals. Using structured programming as a guide, we analyse the BT principles of reactiveness and modularity in a formal framework of action selection. Proceeding from these principles, we review a number of challenging use cases of BTs in the literature, and show that reasoning via these principles leads to compatible solutions. Extending these arguments, we introduce a new class of control architectures we call generalised BTs or $k$-BTs and show how they can extend the applicability of BTs to some of the aforementioned challenging BT use cases while preserving the BT principles.

Jack Henderson, Jochen Trumpf, Mohammad Zamani

We present a new approach to the cooperative localisation problem by applying the theory of minimum energy filtering. We consider the problem of estimating the pose of a group of mobile robots in an environment where robots can perceive fixed landmarks and neighbouring robots as well as share information with others over a communication channel. Whereas the vast majority of the existing literature applies some variant of a Kalman Filter, we derive a set of filter equations for the global state estimate based on the principle of minimum energy filtering. We show how the filter equations can be decoupled and the calculations distributed among the robots in the network without requiring a central processing node. Finally, we provide a demonstration of the filter's performance in simulation.