Performance Of Neural Controller Research Articles

An improved indirect adaptive neural control performance based on MOPSO approach: An experimental validation via a transesterification reactor

This paper proposes an indirect adaptive control method based on recurrent neural networks. To achieve satisfactory closed-loop performances, a neural emulator (NE) and a neural controller (NC) adapting rates are established using the multiobjective particle swarm optimization (MOPSO) algorithm. The proposed MOPSO algorithm has been designed to minimize, simultaneously, two separated objective functions: the emulation and the tracking errors. The proposed approach guarantees that the NE tracks the system dynamics within a short time window. Consequently, it provides for the suggested control structure useful information to synthesize optimal adaptive rates of the NE and NC. To validate the effectiveness of the proposed MOPSO algorithm, a numerical example and an experimental validation on a chemical reactor are proposed.

Closed-Loop Continuous Hand Control via Chronic Recording of Regenerative Peripheral Nerve Interfaces.

Loss of the upper limb imposes a devastating interruption to everyday life. Full restoration of natural arm control requires the ability to simultaneously control multiple degrees of freedom of the prosthetic arm and maintain that control over an extended period of time. Current clinically available myoelectric prostheses do not provide simultaneous control or consistency for transradial amputees. To address this issue, we have implemented a standard Kalman filter for continuous hand control using intramuscular electromyography (EMG) from both regenerative peripheral nerve interfaces (RPNI) and an intact muscle within non-human primates. Seven RPNIs and one intact muscle were implanted with indwelling bipolar intramuscular electrodes in two rhesus macaques. Following recuperations, function-specific EMG signals were recorded and then fed through the Kalman filter during a hand-movement behavioral task to continuously predict the monkey's finger position. We were able to reconstruct continuous finger movement offline with an average correlation of and a root mean squared error (RMSE) of 0.12 between actual and predicted position from two macaques. This finger movement prediction was also performed in real time to enable closed-loop neural control of a virtual hand. Compared with physical hand control, neural control performance was slightly slower but maintained an average target hit success rate of 96.70%. Recalibration longevity measurements maintained consistent average correlation over time but had a significant change in RMSE ( ). Additionally, extracted single units varied in amplitude by a factor of +18.65% and -25.85% compared with its mean. This is the first demonstration of chronic indwelling electrodes being used for continuous position control via the Kalman filter. Combining these analyses with our novel peripheral nerve interface, we believe that this demonstrates an important step in providing patients with more naturalistic control of their prosthetic limbs.

Neural control of finger movement via intracortical brain–machine interface

Objective. Intracortical brain–machine interfaces (BMIs) are a promising source of prosthesis control signals for individuals with severe motor disabilities. Previous BMI studies have primarily focused on predicting and controlling whole-arm movements; precise control of hand kinematics, however, has not been fully demonstrated. Here, we investigate the continuous decoding of precise finger movements in rhesus macaques. Approach. In order to elicit precise and repeatable finger movements, we have developed a novel behavioral task paradigm which requires the subject to acquire virtual fingertip position targets. In the physical control condition, four rhesus macaques performed this task by moving all four fingers together in order to acquire a single target. This movement was equivalent to controlling the aperture of a power grasp. During this task performance, we recorded neural spikes from intracortical electrode arrays in primary motor cortex. Main results. Using a standard Kalman filter, we could reconstruct continuous finger movement offline with an average correlation of ρ = 0.78 between actual and predicted position across four rhesus macaques. For two of the monkeys, this movement prediction was performed in real-time to enable direct brain control of the virtual hand. Compared to physical control, neural control performance was slightly degraded; however, the monkeys were still able to successfully perform the task with an average target acquisition rate of 83.1%. The monkeys’ ability to arbitrarily specify fingertip position was also quantified using an information throughput metric. During brain control task performance, the monkeys achieved an average 1.01 bits s−1 throughput, similar to that achieved in previous studies which decoded upper-arm movements to control computer cursors using a standard Kalman filter. Significance. This is, to our knowledge, the first demonstration of brain control of finger-level fine motor skills. We believe that these results represent an important step towards full and dexterous control of neural prosthetic devices.

A direct and neural controller performance study with beam wandering mitigation control in free space optical link

The beam wander on the detector plane is one of the main causes for major power loss which severely degrades the performance of Free Space Optical (FSO) link. Designing a suitable controller to correct the beam motion at a faster rate to increase the beam stability becomes significant. This paper presents an investigation on the performance of two types of controller designed for aiming a laser beam to be at a particular point under dynamic disturbances. The first design is based on the Taguchi's method (direct controller) while the second is the Artificial Neural Network (ANN) method (neural-controller). These controllers process the beam location information and generate the necessary outputs to mitigate the beam wandering. The pipelined-parallel architecture for both controllers are developed in a Field Programmable Gate Array (FPGA) and installed at the receiver station. Evidence of the suitability and the effectiveness of the proposed controller in terms of prediction exactness, prediction error, reduction of focal point wander, response to impulse and scintillation are provided through experimental results from the FSO link established for the horizontal range of 0.5 km at an altitude of 15.25 m.

Development of a neural controller applied in a 5 DOF robot redundant

In this paper the development of a neural controller implemented in a five Degrees Of Freedom (DOF) redundant robot is presented. The design of the control law considers the robotic system inverse model, including the performance of the actuators for the five joints, obtained through a feedforward neural network with backpropagation learning algorithm. This inverse structure is weighted by desired acceleration and derivative proportional feedback loops to provide the appropriate supply voltage to the servo motors of the robotic manipulator. Tracking tests are performed to a path in Cartesian space using a simulator developed using MatLab/Simulink software tools. It assesses the neural controller performance versus classical computed torque controller, comparing the results of curves in the joint space and Cartesian through RMS errors indices of Cartesian and joint positions.

Indirect adaptive structure for multivariable neural identification and control of a pilot distillation plant

This paper describes the design and implementation of an indirect adaptive controller that uses neural networks both for identification and control of an experimental pilot distillation column containing a mixture of ethanol and water. The MATLAB platform is applied both for the neural identification and control of the distillation plant using the Levenberg–Marquardt approach, enabling also optimal input/output net configuration. The neural controller performance has been analyzed and illustrated via experimental tests on the pilot distillation column monitored under the LabVIEW platform. Both platforms have been linked together by constituting an integrated process control interface. The obtained experimental results demonstrate the effectiveness of the neural indirect adaptive control scheme as compared to proportional–integrative–derivative, when real-time multivariable control is demanded, even in presence of disturbances.

Evolution of Neural Controllers for Robot Navigation in Human Environments

Problem statement: In this study, we presented a novel vision-based learning approach for autonomous robot navigation. Approach: In our method, we converted the captured image in a binary one, which after the partition is used as the input of the neural controller. Results: The neural control system, which maps the visual information to motor commands, is evolved online using real robots. Conclusion/Recommendations: We showed that evolved neural networks performed well in indoor human environments. Furthermore, we compared the performance of neural controllers with an algorithmic vision based control method.

A vision-based approach for intelligent robot navigation

In this paper, we present an evolutionary approach for vision-based robot navigation in human environments. In our method, we convert the captured image in a binary one, which after the partition is used as the input of the neural controller. The neural control system, which maps the visual information to motor commands, is evolved online using real robots. We show that evolved neural networks performed well in indoor human environments. Furthermore, we compare the performance of neural controllers with an algorithmic vision-based control method.

Intelligent control of a pH process

AbstractA new strategy, to augment the pH process control is offered in this paper. The intelligent controller proposed herein is based on an inverse neural plant model. An integration term is introduced to improve the pure inverse neural controller performance. This element, adjusted by a fuzzy system with respect to the control error, operates in parallel with the neural controller to ensure offset-free performance, in case of system uncertainties or modelling mismatch. Four fuzzy rules were applied to generate the integrator parameters. Experimental results, carried out under pH control on a laboratory scale set-up, demonstrate the feasibility of the proposed control system.

Direct Adaptive Neural Flight Controller for F-8 Fighter Aircraft

A novel neural network approach based on model-following direct adaptive control system design is proposed to improve damping and also to follow pilot commands accurately. The control law is derived using system theory and an implicit function theorem. A neural network with a linear filter is used to approximate the control law. The neural controller is trained using offline (finite interval of time) and online learning strategies. The neural controller was trained offline using the error signal between the aircraft response and pilot command (reference model). The offline-trained neural controller provides the necessary damping and tracking performances. The neural controller is adapted online for variations in aerodynamic coefficients or control surface deficiencies caused by control surface damage. A discrete time linear dynamic longitudinal model of a high-performance aircraft (F-8) is considered to demonstrate the effectiveness of the proposed control scheme. The performance results of the proposed control scheme are compared with the recently developed dynamic inversion technique and fully tuned radial basis function network. The neural controller performance is also evaluated for steady climb-and-hold maneuver in a nonlinear six-degree-of-freedom aircraft model.

Real-Time Implementation of Model Predictive Neural Controller for Heat Exchanger

The control of heat exchanger is complex due to its nonlinear dynamics, and particularly the variable steady-state gain and time constant with the process fluid. Linear controllers designed based on linear models will be effective only in a small region around the operating point. The artificial neural network (ANN) technique is used to control the temperature of hot fluid flowing in the inner tube of a physical heat exchanger setup. A methodology is proposed for training and prediction of dynamic behaviour of heat exchanger using feed-forward neural network with external recurrent connections. Then a nonlinear predictive control strategy based on identified model is proposed for heat exchanger control. The performances of neural controller are evaluated in real time and results are compared with a localized PID controller.

CONTROL OF pH IN A LABORATORY FERMENTER USING NEURO-FUZZY TECHNIQUE

This contribution offers a new strategy, to augment the pH process control in a laboratory scale fermenter, based on inverse neural plant model. An integration term is introduced to improve the pure neural controller performance. This element, adjusted by a fuzzy system with respect to the control error, operates in parallel with neural controller to ensure offset-free performance in case of system uncertainties or model mismatch. Four fuzzy rules were applied to generate the integrator parameters. Experimental results demonstrate the usefulness of the fuzzy integrating term and the robustness of the proposed control system.

Application of recurrent neural networks in batch reactors: Part II: Nonlinear inverse and predictive control of the heat transfer fluid temperature

Although nonlinear inverse and predictive control techniques based on artificial neural networks have been extensively applied to nonlinear systems, their use in real time applications is generally limited. In this paper neural inverse and predictive control systems have been applied to the real-time control of the heat transfer fluid temperature in a pilot chemical reactor. The training of the inverse control system is carried out using both generalised and specialised learning. This allows the preparation of weights of the controller acting in real-time and appropriate performances of inverse neural controller can be achieved. The predictive control system makes use of a neural network to calculate the control action. Thus, the problems related to the high computational effort involved in nonlinear model-predictive control systems are reduced. The performance of the neural controllers is compared against the self-tuning PID controller currently installed in the plant. The results show that neural-based controllers improve the performance of the real plant.

Development of adaptive neural networks for flexible control of batch processes

This paper deals with the application of a neural controller for temperature control of a batch reactor. The term “neural controller” is used to refer to a multilayer neural network which computes the control values to be applied to the process. We present the design and the development of the neural network: architecture, learning database and learning procedure. In a first step, the learning phase consists in teaching the neural network to map the dynamics of a classical adaptive controller (generalized predictive control with double model reference) implemented on the process. Although the neural controller performance is good for operating conditions included in the learning set (interpolation), it exhibits limitations on extrapolation. In this work, two methods for the on-line adaptation of the network's weights are developed: one of them is the “specialized” learning technique, whereas the other uses another neural network in order to model the reactor dynamics. Several results are shown and prove the good capacities of neural networks for controlling batch processes.

A knowledge-based approach to improve neural network control design

Control engineers have been investigating and developing different on-line adaptation schemes to fine-tune performance of controllers after off-line design. Artificial neural network technology has shown satisfactory results for many control applications. However, certain types of nonlinear problems are difficult for the neural controller to learn by conventional on-line adaptation schemes, while the nonlinear system can be effectively controlled by incorporating heuristics knowledge. This paper presents an effective approach to incorporate heuristics control knowledge into a neural controller by off-line pre-training, then fine-tune the neural controller performance further by on-line adaptation. Experimental results on a servomotor system with significant nonlinear friction characteristics are used to demonstrate the effectiveness of the design approach.