| Paper basis weight is defined as the weight of paper per unit area which is an important indicator affecting the quality and cost of paper.Basis weight control consists of Machine Direction(MD)basis weight control and Cross Directional(CD)basis weight control,with the trend towards lower basis weight of paper and paperboard as well as the increasing consumer demand for paper quality,which is difficult to meet with MD basis weight control only.Although some developments have been made in the field of CD basis weight control,the control difficulties caused by non-squared high-dimensional,strong coupling and large time-delay characteristics haven’t been well solved.The thesis,which is funded by the National Natural Science Foundation of China(Project No.62073206),aiming at the above control difficulties,theory and applied technology base research has been conducted around strategies of CD basis weight control based on hydraulic headbox with dilution water.The main work of the thesis can be summarised in the following four areas.I.General control scheme design for CD basis weight.Firstly,the structure of the hydraulic headbox with dilution water is partitioned,while the functions and working mechanism of each part are explained;the traditional lip adjustment method is compared with the dilution water adjustment method,so that the advantages of using the dilution water adjustment method for CD basis weight control are derived;Secondly,the main control difficulties of the dilution water adjustment method for CD basis weight control are analysed,and the specific problems that need to be solved are clarified;Finally,the general control scheme of the CD basis weight control system is designed based on the control difficulties obtained from the analysis.II.Research on online feedforward decoupling algorithms for CD basis weight based on ELMAN neural network.Aiming at the non-square,high-dimensional strong coupling problem of CD basis weight control,an online feedforward decoupling algorithm based on ELMAN neural network is proposed.Firstly,the system is reduced by using the idea of "region transformation point" and introducing a matrix of prior transformation factors;then,the ELMAN neural network is combined with feedforward decoupling for the squared system,and on this basis,a reference model is introduced to adjust the network parameters online according to the error between the output of the reference model and the output of the system.The system coupling is compensated in real time to achieve online decoupling of the system.Simulations comparing the algorithm in the thesis with BP neural network decoupling algorithm and adjoint matrix decoupling algorithm show that the algorithm in the thesis is less dependent on the exact coupling model of the system and has a higher decoupling accuracy.Ⅲ.Research on control strategies for CD basis weight based on fuzzy selftuning.Aiming at the problem of large time-delay in the CD basis weight control,a fuzzy self-tuning PID Smith predictive control strategy based on chaos optimization(COFuzzyPID-FSmith)is proposed for the single-loop group of the decoupled system using the idea of " distributed control".Firstly,a feedback filter is introduced to improve the traditional Smith predictor(called FSmith),so that the model mismatch error is weakened by the filter and then fed back;then,considering that the fuzzy control is not highly dependent on the model,it is combined with a PID controller as the main controller of the improved Smith predictor,and chaos optimization is used to optimize the parameters of the main controller.Simulations comparing the control strategy in this thesis with PID,PIDSmith and FuzzyPID-FSmith control strategies show that the control strategy in the thesis has the shorter response time,and the better performance in terms of anti-interference and robustness.Meeting the requirements of CD basis weight control system for fast control in real time with large time-delay.IV.Design and implementation of CD basis weight control system.The general structure of the system is designed according to the general scheme of the CD basis weight control system,laying the foundation for the implementation of the decoupling and time delay control strategies proposed in this thesis.The thesis uses OPC technology for data exchange,adopts S7-400 as the hardware control body,and combines Step7 and WinCC to implement the control strategies.Finally practical engineering application shows that the error of the CD basis weight can be controlled at 1.86%,which achieved a good control effect.In summary,the problem of uniform distribution of CD basis weight is solved by control means in this thesis,focusing on the research of non-square highdimensional system decoupling and large time-delay control strategies,the control strategies designed in this thesis are simulated and applied in practical engineering applications.The results show that the research achievements of the thesis meet the index requirements of the error of the CD basis weight,providing theoretical and technical references for paper manufacturers to carry out CD basis weight control. |
PDF Full Text Request