Design and Optimization of a Neuro-Fuzzy System for the Control of an Electromechanical Plant

One characteristic of neuro-fuzzy systems is the possibility of incorporating preliminary information in their structure as well as being able to establish an initial configuration to carry out the training. In this regard, the strategy to establish the configuration of the fuzzy system is a relevant aspect. This document displays the design and implementation of a neuro-fuzzy controller based on Boolean relations to regulate the angular position in an electromechanical plant, composed by a motor coupled to inertia with friction (a widely studied plant that serves to show the control system design process). The structure of fuzzy systems based on Boolean relations considers the operation of sensors and actuators present in the control system. In this way, the initial configuration of fuzzy controller can be determined. In order to perform the optimization of the neuro-fuzzy controller, the continuous plant model is converted to discrete time to be included in the closed-loop controller training equations. For the design process, first the optimization of a Proportional Integral (PI) linear controller is carried out. Thus, linear controller parameters are employed to establish the structure and initial configuration of the neuro-fuzzy controller. The optimization process also includes weighting factors for error and control action in such a way that allows having different system responses. Considering the structure of the control system, the optimization algorithm (training algorithm) employed is dynamic back propagation. The results via simulations show that optimization is achieved in the linear and neuro-fuzzy controllers using different weighting values for the error signal and control action. It is also observed that the proposed control strategy allows disturbance rejection.

Time-Optimal Gathering under Limited Visibility with One-Axis Agreement

We consider the distributed setting of N autonomous mobile robots that operate in Look-Compute-Move (LCM) cycles following the well-celebrated classic oblivious robots model. We study the fundamental problem of gathering N autonomous robots on a plane, which requires all robots to meet at a single point (or to position within a small area) that is not known beforehand. We consider limited visibility under which robots are only able to see other robots up to a constant Euclidean distance and focus on the time complexity of gathering by robots under limited visibility. There exists an O(DG) time algorithm for this problem in the fully synchronous setting, assuming that the robots agree on one coordinate axis (say north), where DG is the diameter of the visibility graph of the initial configuration. In this article, we provide the first O(DE) time algorithm for this problem in the asynchronous setting under the same assumption of robots’ agreement with one coordinate axis, where DE is the Euclidean distance between farthest-pair of robots in the initial configuration. The runtime of our algorithm is a significant improvement since for any initial configuration of N≥1 robots, DE≤DG, and there exist initial configurations for which DG can be quadratic on DE, i.e., DG=Θ(DE2). Moreover, our algorithm is asymptotically time-optimal since the trivial time lower bound for this problem is Ω(DE).

A mechanobiological model for tumor spheroids evolution: application to glioblastoma

Spheroids are in vitro spherical structures of cell aggregates, eventually cultured within a hydrogel matrix, that are used, among other applications, as a technological platform to investigate tumor formation and evolution. Several interesting features can be replicated using this methodology, such as cell communication mechanisms, the effect of gradients of nutrients, or the creation of realistic 3D biological structures. In this paper, we propose a continuum mechanobiological model which accounts for the most relevant phenomena that take place in tumor spheroids evolution under in vitro suspension, namely, nutrients diffusion in the spheroid, kinetics of cellular growth and death, and mechanical interactions among the cells. The model is qualitatively validated, after calibration of the model parameters, versus in vitro experiments of spheroids of different glioblastoma cell lines. This preliminary validation allowed us to conclude that glioblastoma tumor spheroids evolution is mainly driven by mechanical interactions of the cell aggregate and the dynamical evolution of the cell population. In particular, it is concluded that our model is able to explain quite different setups, such as spheroids growth (up to six times the initial configuration for U-87 MG cell line) or shrinking (almost half of the initial configuration for U-251 MG cell line); as the result of the mechanical interplay of cells driven by cellular evolution. Indeed, the main contribution of this work is to link the spheroid evolution with the mechanical activity of cells, coupled with nutrient consumption and the subsequent cell dynamics. All this information can be used to further investigate mechanistic effects in the evolution of tumors and their role in cancer disease.

Non-inverse kinematics of free-floating space robot based on motion planning of sampling

Motion planning is one of the fundamental technologies for robots to achieve autonomy. Free-floating space robots composed manipulators and base satellite that do not actively control its position and attitude has nonholonomic characteristics, and there is a first-order differential relationship between its joint angle and the base attitude. In addition, the planning framework which first converts the goal end-effector pose to its corresponding target configuration, and then plan the trajectory from the initial configuration to the goal configuration still has the following problems: the goal configuration and the initial configuration may not be in the same connected domain. Based on the RRT framework, the motion planning of a free-floating space robot from the initial configuration to the goal end-effector pose is studied. In the algorithm design, in order to deal with the differential constraints of the free-floating space robot, and the requirement that the attitude disturbance of its base cannot exceed its limit, a control-based local planner for random configuration guiding growth of the tree and a control-based local planner for goal end-effector pose guiding growth of the tree that can adjust the attitude of the base when necessary are proposed. The former can ensure the effective exploration of the configuration space, and the latter can avoid the occurrence of singularity while ensuring that the algorithm converges quickly and the base attitude disturbance meets the constraints. The present algorithm does not need to solve the inverse kinematics, can successfully complete the planning task, and ensure that the base attitude disturbance meets the requirements. The simulation verifies the effectiveness of the algorithm.

Effect of Initial Configuration of Weights on Training and Function of Artificial Neural Networks

The function and performance of neural networks are largely determined by the evolution of their weights and biases in the process of training, starting from the initial configuration of these parameters to one of the local minima of the loss function. We perform the quantitative statistical characterization of the deviation of the weights of two-hidden-layer feedforward ReLU networks of various sizes trained via Stochastic Gradient Descent (SGD) from their initial random configuration. We compare the evolution of the distribution function of this deviation with the evolution of the loss during training. We observed that successful training via SGD leaves the network in the close neighborhood of the initial configuration of its weights. For each initial weight of a link we measured the distribution function of the deviation from this value after training and found how the moments of this distribution and its peak depend on the initial weight. We explored the evolution of these deviations during training and observed an abrupt increase within the overfitting region. This jump occurs simultaneously with a similarly abrupt increase recorded in the evolution of the loss function. Our results suggest that SGD’s ability to efficiently find local minima is restricted to the vicinity of the random initial configuration of weights.

Connect Multi-Agent Path Finding: Generation and Visualization

We present a generic tool to visualize missions of the Connected Multi-Agent Path Finding (CMAPF) problem. This problem is a variant of MAPF which requires a group of agents to navigate from an initial configuration to a goal configuration while maintaining connection. The user can create an instance of CMAPF and can play the generated plan. Any algorithm for CMAPF can be plugged into the tool.

Elastostatic modeling of multi-link flexible manipulator based on 2D dual-triangle tensegrity mechanism

The paper deals with the elastostatic modeling of multi-link flexible manipulator based on dual-triangle tensegrity mechanism and its non-linear behavior under external loading. The main attention is paid to the static equilibriums and the manipulator stiffness behavior under the loading for the arbitrary initial configuration. It was proved that there is a quasi-buckling phenomenon for this manipulator while the external loading is increasing. In the neighborhood of these configurations, the manipulator behavior was analyzed using the enhanced Virtual Joint Method (VJM). A relevant simulation study confirmed the obtained theoretical results.

Self-Stabilizing Distributed Algorithms by Gellular Automata

Gellular automata are cellular automata with the properties of asynchrony, Boolean totality and noncamouflage. In distributed computing, it is essential to determine whether problems can be solved by self-stable gellular automata. From any initial configuration, self-stable gellular automata converge to desired configurations, as self-stability implies the ability to recover from temporary malfunctions in transitions or states. This paper shows that three typical problems in distributed computing, namely, solving a maze, distance-2 coloring and spanning tree construction, can be solved with self-stable gellular automata.

A jamming plane of sphere packings

The concept of jamming has attracted great research interest due to its broad relevance in soft-matter, such as liquids, glasses, colloids, foams, and granular materials, and its deep connection to sphere packing and optimization problems. Here, we show that the domain of amorphous jammed states of frictionless spheres can be significantly extended, from the well-known jamming-point at a fixed density, to a jamming-plane that spans the density and shear strain axes. We explore the jamming-plane, via athermal and thermal simulations of compression and shear jamming, with initial equilibrium configurations prepared by an efficient swap algorithm. The jamming-plane can be divided into reversible-jamming and irreversible-jamming regimes, based on the reversibility of the route from the initial configuration to jamming. Our results suggest that the irreversible-jamming behavior reflects an escape from the metastable glass basin to which the initial configuration belongs to or the absence of such basins. All jammed states, either compression- or shear-jammed, are isostatic and exhibit jamming criticality of the same universality class. However, the anisotropy of contact networks nontrivially depends on the jamming density and strain. Among all state points on the jamming-plane, the jamming-point is a unique one with the minimum jamming density and the maximum randomness. For crystalline packings, the jamming-plane shrinks into a single shear jamming-line that is independent of initial configurations. Our study paves the way for solving the long-standing random close-packing problem and provides a more complete framework to understand jamming.