Fuzzy model predictive control for small-scale biomass combustion furnaces

Böhler, Lukas; Krail, Jürgen; Görtler, Gregor; Kozek, Martin

doi:10.1016/j.apenergy.2020.115339

Cited by 23 publications

(8 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, a decrease in the activation energy of microwave pyrolysis of corn feed, lignin, cellulose and hemicellulose by about 15.2 %, 17.3 %, 19.9 % and 11.8 %, respectively, was achieved. Other method of taking into account the nonlinearities of the pyrolysis reactor is the use of a predictive model, as shown in [3]. In this work, using experimental data, a fuzzy model of the states of a pyrolysis reactor was constructed, which made it possible to use adaptive control over the state vector.…”

Section: Research Of Existing Solutions To the Problemmentioning

confidence: 99%

Development of control system for waste pyrolysis unit of agricultural complex with the application of fuzzy logic

Maksymenko

2021

TAPR

View full text Add to dashboard Cite

The object of research is the control system for the pyrolysis reactor of agricultural waste (plant biomass). The subject of research is the stability and the value of the calorific value of synthesis gas formed by pyrolysis of plant biomass. The biggest problem of the technological object (the pyrolysis reactor of agricultural waste) is the high sensitivity of the heating value of synthesis gas to disturbances in the composition of plant biomass. This sensitivity is expressed as a square law of the amount of oxidant required to achieve a high calorific value. Another problem is the deviation of certain time constants of the control object, caused by changes in the chemical composition of the plant biomass. The built control system provides a high calorific value of the generated syngas by determining the composition of the waste, pyrolysis by determining the composition of the generated syngas in a separate isoenthalpic device, and stabilizes it. Information on the composition of raw materials allows to calculate the optimal parameters for the pyrolysis process, and, accordingly, update the controller's task. This information also makes it possible to compensate for changes in the time constants of the control object caused by changes in the chemical composition of raw materials, which made it possible to achieve a high robustness of the system. Compensation for these changes was carried out by training a regression polynomial. The training was carried out on test sets of time constant deviations. The resulting polynomials were used for convolution with membership functions of a fuzzy controller. Such a convolution made it possible to obtain the following membership functions that ensure compliance with the control quality parameters close to those obtained without deviations in the time constants. Simulation of the constructed control system showed a significantly reduced sensitivity of the calorific value to the composition of raw materials, and also revealed a low sensitivity of the control quality from the deviations of the time constants of the control object caused by disturbances in the chemical composition of the waste. The method by which the control system for the pyrolysis reactor was built differs from the existing ones in that the use of information on the composition of the pyrolyzed substance is used to accurately calculate the optimal values of the pyrolysis parameters, as well as to mutate the membership functions of the fuzzy controller. The method can be used in other similar systems designed for the pyrolysis of organic substances in order to expand their scope. In particular, for the integration of such systems into technological objects, they are more sensitive to deviations in the calorific value of the gas used as fuel.

show abstract

Section: Research Of Existing Solutions To the Problemmentioning

confidence: 99%

Development of control system for waste pyrolysis unit of agricultural complex with the application of fuzzy logic

Maksymenko

2021

TAPR

View full text Add to dashboard Cite

show abstract

“…In Fuzzy system approach, almost all studies have been carried out based on Takagi-Sugeno (TS) models in which the consequent part of rules is mostly of linear-type, as in (Karer et al, 2008) and (Palm and Driankov 1998). In (Böhler et al, 2020), a small-scale grate furnace is modeled by a newly derived biomass combustion model in which the nonlinear process model is split into a number of linearized submodels. Then, several local linear controllers are designed for these submodels using a gap metric.…”

Section: Introductionmentioning

confidence: 99%

A g-Normal fuzzy relational hybrid model based control design for hybrid systems with continuous-valued inputs

Bakhoda

Menhaj

Ashtiani

2022

Journal of Vibration and Control

View full text Add to dashboard Cite

In this paper, the indirect adaptive control of a special class of hybrid systems is proposed. The proposed control scheme is based on g-normal fuzzy relational hybrid models (g-Normal FRHMs) in which the model of both plant and controller are proposed to be modeled by FRHM structure. There is an assumption on control action to be of continuous-valued type. In order to design a control scheme for both regulation and reference tracking, an additional simplifying assumption is considered. Indeed, when the tuning mechanism is extracted, it is assumed that the defuzzifier of the controller’s model and the fuzzifier of the plant’s model are approximately inverse of each other and can be simplified. The proposed control approach is then applied to a DC-DC buck converter as our case study. The simulation results show that the proposed adaptive control based on g-Normal FRHM can regulate and also track the reference input in the presence of both load and supply voltage disturbance.

show abstract

“…The receding horizona quantity of computing steps of the future model states based on the present and past system conditionsis an additional adjustment parameter of a final control law. Some researchers combine MPC with different control approaches like fuzzy-logic [10], artificial neural networks [11] or it can be used for optimal set points or tracking trajectories [12] for local control loops with familiar controllers. Hierarchical structures of MPC because of complexity of closed solutions are proposed in [13].…”

Section: Introductionmentioning

confidence: 99%

Model Predictive Control Toolbox Design for Nonstationary Process

Stepanets

Mariiash

2021

KPI SN

View full text Add to dashboard Cite

Background. Model predictive control (MPC) approach is the basic feedback scheme, combined with high adaptive properties, which determines its successful use in the practice of design and operation of control systems. These advantages allow managing multidimensional objects with a complex structure, including nonlinearity, optimizing processes in real time within the constraints on controlled and managed variables, taking into account uncertainties in the task of objects and perturbations. Objective. The purpose of the paper is to design and analyse control system of carbon monoxide oxidation in the convector cavity based on MPC with linear-quadratic cost functional with constraint. Methods. The design of MPC is based on mathematical model of an object (relatively simple). At the current step, the prediction of object dynamic response on some final period of time (prediction horizon) is carried out; control optimization is performed, the purpose of which is to approximate the control variables of the prediction model to the corresponding setpoint on the predict horizon. The found optimal control is applied and measurement of an actual state of object at the end of a step is carried out. The prediction horizon is shifted one step further, and this algorithm are repeated. Results. The results of modeling the automatic control system show that the MPC approach provides maintenance of carbon dioxide content when changing oxygen consumption and overshoot caused by introduction bulk does not exceed 0.6 % that meets the technological requirements of the process. Conclusions. A fuse of the MPC and the quadratic functional given the constraints on the input signals is proposed. The problems of control degree of carbon oxidation in the convector cavity include non-stationarity, so the use of classical control methods is difficult. The MPC approach minimizes the cost function that characterizes the quality of the process. The predicted behaviour of a dynamic system will usually differ from its actual motion. The obtained quadratic functional is optimized to find the optimal control of degree of CO oxidation to CO2.

show abstract

Fuzzy model predictive control for small-scale biomass combustion furnaces

Cited by 23 publications

References 32 publications

Development of control system for waste pyrolysis unit of agricultural complex with the application of fuzzy logic

Development of control system for waste pyrolysis unit of agricultural complex with the application of fuzzy logic

A g-Normal fuzzy relational hybrid model based control design for hybrid systems with continuous-valued inputs

Model Predictive Control Toolbox Design for Nonstationary Process

Contact Info

Product

Resources

About