Fuzzy Logic Based Control Strategies for an Electromagnetic Actuated Sewing Machine Presser Foot
An adaptive method was used for all the presented controllers to set up a new reference value whenever the error or force output exceeds a defined range. The method allows adapting different fabrics and a different number of plies. The problem is that sewing defects like folds, may be incorrectly interpreted and a new reference value may be set, an undesired situation.. Therefore new techniques will be considered in the future to avoid adaptation to these sewing defects. VS Sewing Machines The definition of references is being implemented by running a specific sewing test before the operation, in order to determine the reference values based on an anti-bouncing condition.
Fuzzy Logic Controller
In order to define the controller parameters, some tests were carried out varying the sewing speed and the force output. The displacement of ihe presser foot was measured during these tests, for a range of sewing speeds and presser-foot forces. In Fig. 2 an example of the characterisation of two plies of an interlock fabric is depicted. In this figure it is possible to observe how the displacement varies with force and speed. In terms of controller design, and based on the result of practical sewing tests, the controller should maintain a displacement range of about 0.05″ around a Qhrc #educt 5 dm pre-defined average value. The correspondent force range lies between 17.4 to 26.XN for a speed variation from 900 to 4700 stitches per minute (spm).
Another interesting fact is that the sewing speed needs also to be taken into account, but seems to be less important. With the help of these graphs for the defined fabrics, the error range was defined from -0.05 to 0.05″ and the force range from 17.4N to 26.8N. This force range seems to manage the defined displacement error over the used speed range. Based on these principles the Fuzzy Logic Controller was designed. https://vssewingmachine.in/merritt-sewing-machine-price-in-chennai/ Due to the preliminary studies concerning the setting of the displacement reference, it was decided to keep the adaptive method of the reference based on the error range. The Fuzzy Logic Controller uses three inputs. One of the thee inputs is the error based on the reference and the current displacement. In addition to the error, the speed and the current displacement were used as input variables to the Fuzzy Logic Controller to ensure an adaptation on the speed and on the number of plies. The control scheme is depicted in Fig. 3. With these three input variables and the force as output, the member-functions were defined (Fig, 4 to Fig. 7). In Fig. 4 the speed was defined in a range of 0 to 47OOspm. The member-functions are divided into low (0 to 2350spm), mid (0 to 4700spm) and high speed (2350 to 4700spm). In Fig. 5 the error is depicted as second antecedent. It was fixed from -0.05 to 0.05″ and divided into memberfunctions of negative, zero and positive.
As the third antecedent the displacement was used, in order to differentiate between different types of fabrics as well as different number of plies to adapt the force. For this purpose, ranges were defined from 0.8 to 0.98mm for two plies of riblxl fabric, from 0.93 to 1.12″ for two plies of interlock fabric, from 1.35 to 1.55″ for four plies of riblxl and from 1.55 to 1.80″ for four plies of interlock fabric, as the member-fimctions depicted in Fig. 6. The output force variable was fixed to a range from 17.4 to 26.8N based on the experience gained on the sewing process.
The objective was the development of fuzzy logic based controller strategies. During this work, a Fuzzy Logic Controller was developed and tested. Besides the Fuzzy Logic Controller, another control strategy was implemented based on a parallel structure combing the Fuzzy Logic Controller with a PID-structure as an auxiliary method to correct only small errors. In order to evaluate the Fuzzy Logic Controller and the combined controller, a comparison with a previously developed reference adapting P-controller was made. This simple P-controller just uses the error as feedback variable, with a constant force offset ’. The comparison of the adapting P-controller and the adapting Fuzzy Logic Controller revealed that the Fuzzy Controller offers more possibilities of specific adaptation of the sewing parameters like speed, number of plies and type of fabric. Nevertheless, this controller requires more testing and verification while establishing the rules for a desirable operation a fine-tuning of the force, depending on the error, demands a close verification. The Fuzzy Logic Controller can be improved by additional member-functions for the error, which will define its behaviour if the error exceeds the defined range. This behaviour will cover the transition zones while changing the number of plies. This is, for example, one advantage compared to the Pcontroller, which just gives maximum or minimum force output at this transition phase. An adaptation to sewing speed was implemented as a property of the fuzzy controller. The P-controller is indirectly influenced by sewing speed: if a higher sewing speed produces larger errors, the p-control will react to them and thus be indirectly influenced by speed. That’s why the results of the P-controIler and Fuzzy Logic Controller are similar for the tested sewing speeds. The used Fuzzy Logic controller adapts also to the number of plies and the type of fabric based on the displacement. This feature was not possible with a P- or PID-controller. Therefore an adaptation of the force ranges is needed for the P-controlter.
The disadvantage of not adapting the number of plies or type of fabric or speed, led to the development of a parallel controller structure. This parallel controller structure was made of a Fuzzy Logic Controller and a PID-Controller. The task of the Fuq Logic Controller was set-up a force offset depending on the sewing parameters. The PIDController served to eliminate the error between the reference value and the actual displacement measured. The rested controller structure seems to be easier to handle, because it does not need a careful calibration like the Fuzzy Logic Controller tested before. The recognition of the fabrics and the number of plies depends on the definitions of the member functions in the Fuzzy Logic Controller. This definition is fixed’in the controller and therefore not applicable for new types of fabrics.