Near optimal manufacturing flow controller design

Flow control of flexible manufacturing systems (FMSs) addresses an important real-time scheduling requirement of modern manufacturing facilities, which are prone to failures and other controllable or stochastic discrete events affecting production capacity, such as change of setup and maintenance scheduling. Flow controllers are useful both in the coordination of interconnected flexible manufacturing cells through distributed scheduling policies and in the hierarchical decomposition of the planning and scheduling problem of complex manufacturing systems. Optimal flow-control policies are hedging-point policies characterized by a generally intractable system of stochastic partial differential equations. This article proposes a near optimal controller whose design is computationally feasible for realistic-size systems. The design exploits a decomposition of the multiple-part-type problem to many analytically tractable one-part-type problems. The decomposition is achieved by replacing the polyhedra production capacity sets with inscribed hypercubes. Stationary marginal densities of state variables are computed iteratively for successive trial controller designs until the best inscribed hypercubes and the associated optimal hedging points are determined. Computational results are presented for an illustrative example of a failureprone FMS.     

An architecture for a shop-floor controller using colored Petri nets

In a dynamic and flexible manufacturing environment, a shop-floor controller must be designed so that it automatically (or with minimum human intervention) and quickly responds to the changes (e.g., in part type or part routing) in the system. Such a performance may be achieved provided that the controller is simple and sufficiently general in its scope of application. In this article, we present an architecture for such a shop-floor controller. The architecture is based on colored Petri nets with ordered colored sets and structured input and output functions.     

New combinatorial clustering methods

Sixteen clustering methods are compatible with the general recurrence equation of combinatorial SAHN (sequential, agglomerative, hierarchical and nonoverlapping) classificatory strategies. These are subdivided into two classes: the d-SAHN methods seek for minimal between-cluster distances the h-SAHN strategies for maximal within-cluster homogeneity. The parameters and some basic features of all combinatorial methods are listed to allow comparisons between these two families of clustering procedures. Interest is centred on the h-SAHN techniques; the derivation of updating parameters is presented and the monotonicity properties are examined. Three new strategies are described, a weighted and an unweighted variant of the minimization of the increase of average distance within clusters and a homogeneity-optimizing flexible method. The performance of d- and h-SAHN techniques is compared using field data from the rock grassland communities of the Sashegy Nature Reserve, Budapest, Hungary.Abbreviations CP = Closest pair - RNN = Reciprocal nearest neighbor - SAHN = Sequential, agglomerative, hierarchical and nonoverlapping     

Novel recurrent neural network for modelling biological networks: Oscillatory p53 interaction dynamics

Understanding the control of cellular networks consisting of gene and protein interactions and their emergent properties is a central activity of Systems Biology research. For this, continuous, discrete, hybrid, and stochastic methods have been proposed. Currently, the most common approach to modelling accurate temporal dynamics of networks is ordinary differential equations (ODE). However, critical limitations of ODE models are difficulty in kinetic parameter estimation and numerical solution of a large number of equations, making them more suited to smaller systems. In this article, we introduce a novel recurrent artificial neural network (RNN) that addresses above limitations and produces a continuous model that easily estimates parameters from data, can handle a large number of molecular interactions and quantifies temporal dynamics and emergent systems properties. This RNN is based on a system of ODEs representing molecular interactions in a signalling network. Each neuron represents concentration change of one molecule represented by an ODE. Weights of the RNN correspond to kinetic parameters in the system and can be adjusted incrementally during network training. The method is applied to the p53-Mdm2 oscillation system – a crucial component of the DNA damage response pathways activated by a damage signal. Simulation results indicate that the proposed RNN can successfully represent the behaviour of the p53-Mdm2 oscillation system and solve the parameter estimation problem with high accuracy. Furthermore, we presented a modified form of the RNN that estimates parameters and captures systems dynamics from sparse data collected over relatively large time steps. We also investigate the robustness of the p53-Mdm2 system using the trained RNN under various levels of parameter perturbation to gain a greater understanding of the control of the p53-Mdm2 system. Its outcomes on robustness are consistent with the current biological knowledge of this system. As more quantitative data become available on individual proteins, the RNN would be able to refine parameter estimation and mapping of temporal dynamics of individual signalling molecules as well as signalling networks as a system. Moreover, RNN can be used to modularise large signalling networks.     

Reinforcement learning with internal expectation in the random neural networks for cascaded decisions

The reinforcement learning scheme proposed in Halici (J. Biosystems 40 (1997) 83) for the random neural network (RNN) (Neural Computation 1 (1989) 502) is based on reward and performs well for stationary environments. However, when the environment is not stationary it suffers from getting stuck to the previously learned action and extinction is not possible. To overcome the problem, the reinforcement scheme is extended in Halici (Eur. J. Oper. Res., 126(2000) 288) by introducing a new weight update rule (E-rule) which takes into consideration the internal expectation of reinforcement. Although the E-rule is proposed for the RNN, it can be used for training learning automata or other intelligent systems based on reinforcement learning. This paper looks into the behavior of the learning scheme with internal expectation for the environments where the reinforcement is obtained after a sequence of cascaded decisions. The simulation results have shown that the RNN learns well and extinction is possible even for the cases with several decision steps and with hundreds of possible decision paths.     

Multiple-Objective Scheduling for the Hierarchical Control of Flexible Manufacturing Systems

This paper presents a hierarchical approach to scheduling flexible manufacturing systems (FMSs) that pursues multiple performance objectives and considers the process flexibility of incorporating alternative process plans and resources for the required operations. The scheduling problem is solved at two levels: the shop level and the manufacturing system level. The shop level controller employs a combined priority index developed in this research to rank shop production orders in meeting multiple scheduling objectives. To overcome dimensional complexity and keep a low level of work-in-process inventory, the shop controller first selects up to three production orders with the highest ranking as candidates and generates all possible release sequences for them, with or without multitasking. These sequences are conveyed to the manufacturing system controller, who then performs detailed scheduling of the machines in the FMS using a fixed priority heuristic for routing parts of multiple types while considering alternative process plans and resources for the operations. The FMS controller provides feedback to the shop controller with a set of suggested detailed schedules and projected order completion times. On receiving these results, the shop controller further evaluates each candidate schedule using a multiple-objective function and selects the best schedule for execution. This allows multiple performance objectives of an FMS to be achieved by the integrated hierarchical scheduling approach.     

Assessing the structural complexity of manufacturing systems configurations

Modern manufacturing systems are increasingly required to be flexible and adaptable to changing market demands, which adds to their structural and operational complexity. One of the major challenges at the early design stages is to select a manufacturing system configuration that both satisfies the production functional requirements and is easy to operate and manage. A new metric for assessing the structural complexity of manufacturing system configurations is presented in this paper. The proposed complexity metric incorporates the quantity of information using an entropy approach. It accounts for the complexity inherent in the various modules in the manufacturing system through the use of an index derived from a newly developed manufacturing systems classification code. The code captures the effect of various component types and technologies used in a manufacturing system on the system’s structural complexity. The presented metric would be helpful in selecting the least complex manufacturing system configuration that meets the requirements. An engine cylinder head production system is used to illustrate the application of the proposed methodology in comparing feasible but different manufacturing system configurations capable of producing the cylinder head based on their structurally inherent complexity.     

Nonlinear system identification and control based on modular neural networks

A new approach for nonlinear system identification and control based on modular neural networks (MNN) is proposed in this paper. The computational complexity of neural identification can be greatly reduced if the whole system is decomposed into several subsystems. This is obtained using a partitioning algorithm. Each local nonlinear model is associated with a nonlinear controller. These are also implemented by neural networks. The switching between the neural controllers is done by a dynamical switcher, also implemented by neural networks, that tracks the different operating points. The proposed multiple modelling and control strategy has been successfully tested on simulated laboratory scale liquid-level system.     

A novel approach for solving constrained nonlinear optimization problems using neurofuzzy systems

A neural network model for solving constrained nonlinear optimization problems with bounded variables is presented in this paper. More specifically, a modified Hopfield network is developed and its internal parameters are computed using the valid-subspace technique. These parameters guarantee the convergence of the network to the equilibrium points. The network is shown to be completely stable and globally convergent to the solutions of constrained nonlinear optimization problems. A fuzzy logic controller is incorporated in the network to minimize convergence time. Simulation results are presented to validate the proposed approach.     

A Supervisory Control Approach for Execution Control of an FMC

This paper presents a generic methodology based on formal language theory for the modeling and control of flexible manufacturing cell (FMC) systems. The motivating idea behind the overall approach stems from the supervisory control theory under the framework of Ramadge and Wonham. Essentially, we characterize the asynchronous and dynamic behavior of an FMC as a regular language and formulate the control logic generation problem as a sublanguage calculation problem, which requires the resulting language to satisfy at least two properties: maximal permissiveness and controllability. Then an algorithm for resolving the problem is presented. Based on the solution of the problem called supervisor, we propose a controller architecture that guarantees coordinated operation of an FMC through the regulation of occurrences of events. An adaptive control policy that regenerates supervisors on changes in task configurations is presented and a dynamic equation that describes the evolution of the control logic along time is derived. Then, we show that the proposed maximally permissive adaptive control policy has a number of preferred properties, including computational efficiency and consistency between the successive supervisors. Finally, a controller for an example FMC is implemented, using the object-oriented software modules. Our procedure has the merit of mathematical soundness, modular design, and systematic implementation.     

Discriminating lysosomal membrane protein types using dynamic neural network

This work presents a dynamic artificial neural network methodology, which classifies the proteins into their classes from their sequences alone: the lysosomal membrane protein classes and the various other membranes protein classes. In this paper, neural networks-based lysosomal-associated membrane protein type prediction system is proposed. Different protein sequence representations are fused to extract the features of a protein sequence, which includes seven feature sets; amino acid (AA) composition, sequence length, hydrophobic group, electronic group, sum of hydrophobicity, R-group, and dipeptide composition. To reduce the dimensionality of the large feature vector, we applied the principal component analysis. The probabilistic neural network, generalized regression neural network, and Elman regression neural network (RNN) are used as classifiers and compared with layer recurrent network (LRN), a dynamic network. The dynamic networks have memory, i.e. its output depends not only on the input but the previous outputs also. Thus, the accuracy of LRN classifier among all other artificial neural networks comes out to be the highest. The overall accuracy of jackknife cross-validation is 93.2% for the data-set. These predicted results suggest that the method can be effectively applied to discriminate lysosomal associated membrane proteins from other membrane proteins (Type-I, Outer membrane proteins, GPI-Anchored) and Globular proteins, and it also indicates that the protein sequence representation can better reflect the core feature of membrane proteins than the classical AA composition.     

Neural solution to the correction of miss distance in gun fire control system

Multilayer perceptrons trained with the backpropagation algorithm are tested in gun fire control system for error correction and are compared to optimal algorithms based on minimum mean square error. The structure of the proposed neural controller is described and performance results are shown.     

A heuristic method to determine the number of pallets in a flexible manufacturing system with several pallet types

This article presents a new heuristic method, called PANORAMA, that enables the determination of an appropriate number of pallets of each type that should circulate in a flexible manufacturing system. Production and return-on-investment constraints are considered. Statistical results derived using data from three existing flexible manufacturing systems, as well as randomly generated data, demonstrate the accuracy of the heuristic proposed.     

A Controller Design Method Based on a Neural Network for an Outdoor Mobile Robot

A wheeled mobile mechanism with a passive and/or active linkage mechanism for travel in rough terrain is developed and evaluated. In our previous research, we developed a switching controller system for wheeled mobile robots in rough terrain. This system consists of two sub-systems: an environment recognition system using a self-organizing map and an adjusted control system using a neural network. In this paper, we propose a new controller design method based on a neural network. The proposed method involves three kinds of controllers: an elementary controller, adjusted controllers, and simplified controllers. In the experiments, our proposed method results in less oscillatory motion in rough terrain and performs better than a well tuned PID controller does.     

Design Guidelines for Deadlock-Handling Strategies in Flexible Manufacturing Systems

Deadlock-free operation of flexible manufacturing systems (FMSs) is an important goal of manufacturing systems control research. In this work, we develop the criteria that real-time FMS deadlock-handling strategies must satisfy. These criteria are based on a digraph representation of the FMS state space. Control policies for deadlock-free operation are characterized as partitioning cuts on this digraph. We call these structural control policies (SCPs) because, to avoid deadlock, they must guarantee certain structural properties of the subdigraph containing the empty state; namely, that it is strongly connected. A policy providing this guarantee is referred to as correct. Furthermore, an SCP must be configurable and scalable; that is, its correctness must not depend on configuration-specific system characteristics and it must remain computationally tractable as the FMS grows in size. Finally, an SCP must be efficient; that is, it must not overly constrain FMS operation. We formally develop and define these criteria, formulate guidelines for developing policies satisfying these criteria, and then provide an example SCP development using these guidelines. Finally, we present an SCP that guarantees deadlock-free buffer space allocation for FMSs with no route restrictions.     

Adaptation supports short-term memory in a visual change detection task

The maintenance of short-term memories is critical for survival in a dynamically changing world. Previous studies suggest that this memory can be stored in the form of persistent neural activity or using a synaptic mechanism, such as with short-term plasticity. Here, we compare the predictions of these two mechanisms to neural and behavioral measurements in a visual change detection task. Mice were trained to respond to changes in a repeated sequence of natural images while neural activity was recorded using two-photon calcium imaging. We also trained two types of artificial neural networks on the same change detection task as the mice. Following fixed pre-processing using a pretrained convolutional neural network, either a recurrent neural network (RNN) or a feedforward neural network with short-term synaptic depression (STPNet) was trained to the same level of performance as the mice. While both networks are able to learn the task, the STPNet model contains units whose activity are more similar to the in vivo data and produces errors which are more similar to the mice. When images are omitted, an unexpected perturbation which was absent during training, mice often do not respond to the omission but are more likely to respond to the subsequent image. Unlike the RNN model, STPNet produces a similar pattern of behavior. These results suggest that simple neural adaptation mechanisms may serve as an important bottom-up memory signal in this task, which can be used by downstream areas in the decision-making process.     

Impacts of nested forward validation techniques on machine learning and regression waste disposal time series models

Dataset partitioning and validation techniques are required in all artificial neural network based waste models. However, there is currently no consensual approach on the validation techniques. This study examines the effects of three time series nested forward validation techniques (rolling origin - RO, rolling window - RW, and growing window - GW) on total municipal waste disposal estimates using recurrent neural network (RNN) models, and benchmarks model performance with respect to multiple linear regression (MLR) models. Validation selection techniques appear important to waste disposal time series model construction and evaluation. Sample size is found as an important factor on model accuracy for both RNN and MLR models. Better performance in Trial RW4 is observed, probably due to a more consistent testing set in 2019. Overall, the MAPE of the waste disposal models ranging from 10.4% to 12.7%. Both GW and RO validation techniques appear appropriate for RNN waste models. However, MLR waste models are more sensitive to the dataset characteristics, and RO validation technique appears more suitable to MLR models. It is found that data characteristics are more important than training period duration. It is recommended data set normality and skewness be examined for waste disposal modeling.     

An annealed chaotic maximum neural network for bipartite subgraph problem

In this paper, based on maximum neural network, we propose a new parallel algorithm that can help the maximum neural network escape from local minima by including a transient chaotic neurodynamics for bipartite subgraph problem. The goal of the bipartite subgraph problem, which is an NP- complete problem, is to remove the minimum number of edges in a given graph such that the remaining graph is a bipartite graph. Lee et al. presented a parallel algorithm using the maximum neural model (winner-take-all neuron model) for this NP- complete problem. The maximum neural model always guarantees a valid solution and greatly reduces the search space without a burden on the parameter-tuning. However, the model has a tendency to converge to a local minimum easily because it is based on the steepest descent method. By adding a negative self-feedback to the maximum neural network, we proposed a new parallel algorithm that introduces richer and more flexible chaotic dynamics and can prevent the network from getting stuck at local minima. After the chaotic dynamics vanishes, the proposed algorithm is then fundamentally reined by the gradient descent dynamics and usually converges to a stable equilibrium point. The proposed algorithm has the advantages of both the maximum neural network and the chaotic neurodynamics. A large number of instances have been simulated to verify the proposed algorithm. The simulation results show that our algorithm finds the optimum or near-optimum solution for the bipartite subgraph problem superior to that of the best existing parallel algorithms.     

Inference of genetic regulatory networks with recurrent neural network models using particle swarm optimization

Genetic regulatory network inference is critically important for revealing fundamental cellular processes, investigating gene functions, and understanding their relations. The availability of time series gene expression data makes it possible to investigate the gene activities of whole genomes, rather than those of only a pair of genes or among several genes. However, current computational methods do not sufficiently consider the temporal behavior of this type of data and lack the capability to capture the complex nonlinear system dynamics. We propose a recurrent neural network (RNN) and particle swarm optimization (PSO) approach to infer genetic regulatory networks from time series gene expression data. Under this framework, gene interaction is explained through a connection weight matrix. Based on the fact that the measured time points are limited and the assumption that the genetic networks are usually sparsely connected, we present a PSO-based search algorithm to unveil potential genetic network constructions that fit well with the time series data and explore possible gene interactions. Furthermore, PSO is used to train the RNN and determine the network parameters. Our approach has been applied to both synthetic and real data sets. The results demonstrate that the RNN/PSO can provide meaningful insights in understanding the nonlinear dynamics of the gene expression time series and revealing potential regulatory interactions between genes.     

Human-like reflex control for an artificial hand

In this paper, we illustrate the low level reflex control used to govern an anthropomorphic artificial hand. The paper develops the position and stiffness control strategy based on dynamic artificial neurons able to simulate the neurons acting in the human reflex control. The controller has a hierarchical structure. At the lowest level there are the receptors able to convert the analogical signal into a neural impulsive signal appropriate to govern the reflex control neurons. Immediately upon it, the artificial motoneurons set the actuators inner pressure to control the finger joint position and moment. Other auxiliary neurons in combination with the motoneurons are able to set the finger stiffness and emulate the inverse myotatic reflex control. Stiffness modulation is important both to save energy during task execution, and to manage objects made of different materials. The inverse myotatic reflex is able to protect the hand from possible harmful external actions. The paper also presents the dynamic model of the joints and of the artificial muscles actuating Blackfingers, our artificial hand. This new type of neural control has been simulated on the Blackfingers model; the results indicate that the developed control is very flexible and efficient for all kind of joints present in the humanoid hand.