Jizhen Liu

Tuning of PID controllers for boiler-turbine units

A simple two-by-two model for a boiler-turbine unit is demonstrated in this paper. The model can capture the essential dynamics of a unit. The design of a coordinated controller is discussed based on this model. A PID control structure is derived, and a tuning procedure is proposed. The examples show that the method is easy to apply and can achieve acceptable performance.

Linear control of a boiler–turbine unit: Analysis and design

Linear control of a boiler-turbine unit is discussed in this paper. Based on the nonlinear model of the unit, this paper analyzes the nonlinearity of the unit, and selects the appropriate operating points so that the linear controller can achieve wide-range performance. Simulation and experimental results at the No. 4 Unit at the Dalate Power Plant show that the linear controller can achieve the desired performance under a specific range of load variations.

Modeling Water Trading under Uncertainty for Supporting Water Resources Management in an Arid Region

AbstractIn this study, a joint-probabilistic interval multistage programming (JIMP) method is developed for planning water resources management under uncertainty. The JIMP method can tackle uncertainties presented in terms of interval parameters in objective function and constraints, in addition to random variables in the left and right-hand sides of constraints. It can also reflect the dynamics in terms of decisions for water resources allocation through transactions at discrete points of a complete scenario set over a multistage context. Moreover, the JIMP method can be used for analyzing various policy scenarios that are associated with different levels of economic consequences when the promised water-allocation targets are violated. The JIMP method is applied to a real case of planning water trading for supporting the regional sustainable development of the Kaidu-Qongque River basin, which is one of the most arid regions of China. Monte Carlo simulation is introduced into the JIMP framework for evalua...

Dynamic model for controller design of condensate throttling systems.

Improving the load adjustment rate of coal-fired power plants in China is very important because of grid load fluctuations and the construction of new large-scale power plants connected to the countrys power grid. In this paper, a new application of condensate throttling system for rapid load adjustment is proposed on the basis of the characteristics of turbine-stored energy. To ensure effective and safe operation of the condensate throttling system, a non-linear control model is derived through reasonable simplifications of fundamental physical laws, and the model parameters are identified using experimental data from a 660 MW supercritical coal-fired power plant. The model outputs are compared with actual measured data for different unit loads. Results show that the established models responses strongly correlate with the actual units responses and can be used for controller design.

Uncertainty analysis for effluent trading planning using a Bayesian estimation-based simulation-optimization modeling approach

In this study, a Bayesian estimation-based simulation-optimization modeling approach (BESMA) is developed for identifying effluent trading strategies. BESMA incorporates nutrient fate modeling with soil and water assessment tool (SWAT), Bayesian estimation, and probabilistic-possibilistic interval programming with fuzzy random coefficients (PPI-FRC) within a general framework. Based on the water quality protocols provided by SWAT, posterior distributions of parameters can be analyzed through Bayesian estimation; stochastic characteristic of nutrient loading can be investigated which provides the inputs for the decision making. PPI-FRC can address multiple uncertainties in the form of intervals with fuzzy random boundaries and the associated system risk through incorporating the concept of possibility and necessity measures. The possibility and necessity measures are suitable for optimistic and pessimistic decision making, respectively. BESMA is applied to a real case of effluent trading planning in the Xiangxihe watershed, China. A number of decision alternatives can be obtained under different trading ratios and treatment rates. The results can not only facilitate identification of optimal effluent-trading schemes, but also gain insight into the effects of trading ratio and treatment rate on decision making. The results also reveal that decision makers preference towards risk would affect decision alternatives on trading scheme as well as system benefit. Compared with the conventional optimization methods, it is proved that BESMA is advantageous in (i) dealing with multiple uncertainties associated with randomness and fuzziness in effluent-trading planning within a multi-source, multi-reach and multi-period context; (ii) reflecting uncertainties existing in nutrient transport behaviors to improve the accuracy in water quality prediction; and (iii) supporting pessimistic and optimistic decision making for effluent trading as well as promoting diversity of decision alternatives.

A modified dynamic matrix control for integrating processes

Dynamic matrix control is a kind of model predictive control technique based on step response model of a process. Conventional DMC is usually not applicable to integrating processes due to an offset in the presence of a sustained step load disturbance. This paper introduces a simple modified DMC algorithm to deal with it. The algorithm proposed is based on the fact that DMC algorithm is equivalent to an IMC structure. Then we can improve conventional DMC using a modified IMC structure to remove the offset. Simulation examples show the effectiveness of control algorithm.

Output tracking control of a nonlinear boiler-turbine unit

The capability to perform fast load changes has been an important issue in the power market. An output tracking control system for improving the load-following capability of boiler-turbine units has therefore been developed. The system is composed of stable inversion and feedback controller. The stable inversion is implemented as a feedforward controller to improve the load-following capability, and the feedback controller is utilized to guarantee the stability and robustness of the whole system. Loop-shaping H/sub /spl infin// method is used to design the feedback controller and the final controller is reduced to a multivariable PI form. The output tracking control system takes account of the multivariable, nonlinear and coupling behavior of boiler-turbine system, and the simulation tests show that the control system works well and applies to a wide range of units.

Identification of Nonlinear System Based on ANFIS with Subtractive Clustering

Subtractive clustering is used to generate an initial T-S fuzzy model with the appropriate rule number and performance index by adjusting the radius of a cluster center. To acquire a T-S fuzzy model with perfect performance, adaptive neuro-fuzzy inference system (ANFIS) is combined to fine tune the premise parameters and consequent parameters by means of a hybrid gradient descent (GD) and least-squares estimation (LSE). A simulation to a dynamic nonlinear system demonstrates the effective of this method

Soft-sensing model of oxygen content based on data fusion

The soft-sensing technique of oxygen content in flue gases based on data fusion is presented according to the high first cost of conventional oxygen content analyzers, their high maintenance expenses and low durability. Through the mechanism analysis and the statistical analysis of a number of data, soft-sensing models of oxygen content and air flow. etc. are set up. Based on multisensor data fusion, more reliable and accurate values of input data are obtained. At last, this soft-sensing model fit well with the practical oxygen content, which is illustrated by the simulations.

Genetic algorithm-based multi-variables nonlinear boiler model identification for 300 MW power unit

A kind of improve genetic algorithm for identifying multi-variables nonlinear boiler model of 300 MW power unit is introduced. In the algorithm, floating-point coding, rank-based selection, elitist reservation and grouping method are used, and the premature convergence is restrained, and the searching ability is improved. The genetic algorithm-based model identification MATLAB program is designed and the model parameters can be gotten with it according to the operating data log files. It is shown by simulation research that the multi-variables nonlinear model can be identified accurately no matter what kind of input signal is used.