PROPOSED CONTROL OF RAW MATERIAL INVENTORY IN CONDITION OF NOT REQUIRED WITH FUZZY MAMDANI METHOD IN CV. PINUS BAG'S SPECIALIST

CV. Pine Bag’s Specialist is a business engaged in the manufacturing of various types of bags. One of the bags made in the form of a backpack. Parachute fabric is the main raw material for making backpacks. Uncertain demand causes a lot of accumulation (over stock) of raw materials in the storage area, so we need a method of supporting raw material inventory control to determine the optimal order. The research objective determines the optimal ordering of raw materials using variable raw material demand, raw material inventory variables and ordering variables in January the first week to March the fourth week. The Fuzzy Mamdani method used in this study because it has a flexible nature and can overcome the problem of uncertainty. The data processing of the Fuzzy Mamdani method carried out in several stages (a) the formation of the Fuzzy set, (b) the application of the implication function, (c) the composition of the rules, (d) Defuzzification. Defuzzification in research uses the centroid method. The results of Fuzzy Mamdani's manual calculation in January of the first week with input of raw material demand of 666 meters and 126 meters of inventory resulted in an optimal prediction of ordering raw materials of 876 meters. Calculations for January the second week to March the fourth week are performed with the help of the Matlab R2013a Fuzzy Toolbox software. The results of prediction data evaluation on the number of raw material orders Fuzzy Mamdani with actual data on the raw material number ordering CV. Pine Bag’s Specialist, it is known that the average absolute error (MAE) is 193.8 meters with an average percentage of absolute error (MAPE) of 22%. So, it is said that the level of accuracy of predictions is reasonable. Future research is expected in the Fuzzy Mamdani method can be used more than two inputs and one output and the addition of linguistic variables. Combine the Fuzzy Mamdani method with other raw material inventory control methods so that the results obtained are getting better.


INTRODUCTION
CV. Pin Bag's Specialist is a company engaged in the manufacturing industry of various types of bags. Backpack is one type of bag that made. The main raw material used in the manufacture of backpacks in the form of cloth called parachute fabric. Companies in making decisions using simple methods manually and based on experience in determining the ordering of raw materials for parachute fabric needed. The simple method used in determining the ordering of parachute fabric raw materials by estimating the amount of demand for parachute p-ISSN : 1412-114X e-ISSN : 2580-5649 http://ojs.pnb.ac.id/index.php/LOGIC fabric raw materials ordered is adjusted to the number of bags production requested by consumers, but the parachute fabric raw material from the rest of the production results is not a consideration factor.
Negative impact if the control of parachute fabric raw materials do not use the right decision support methods in determining the optimal order will cause the accumulation of parachute fabric raw materials (over stock) in the storage area. Parachute fabric raw material that was over stock for the last three months from January 2019 to March 2019 with an average amount of 74.9 meters and a percentage of 12.8% per week. Parachute fabric raw materials that have accumulated in storage for a long time and are not considered causing a decrease in quality that is not suitable for reuse, causing losses for the company.
Based on previous research conducted by Abdurrasyid, et al. [1] resulted in the application of the Fuzzy Mamdani method in determining the predictions of procurement of goods to provide suitable alternatives to meet all user needs, and the application of the Fuzzy Mamdani Method can produce predictions of procurement of goods by looking at the inventory of goods and quantities demand for goods. Research conducted by Rahakbauw, et al [2] resulted in the application of Fuzzy Logic using the Mamdani Method effectively applied in the Matlab software application to assist the company in predicting the determination of the amount of rubber production in liters per day with a percentage of truth value of 87.82706% which means it is very close good. Fuzzy Mamdani method can be used as a decision support in controlling inventory of parachute fabric raw materials based on the amount of raw material demand and inventory of existing raw materials to get optimal ordering predictions in the CV. Pine Bag's Specialist. Proposed inventory control of parachute fabric raw materials using Fuzzy Mamdani method as a decision support is expected to be able to overcome the problem of uncertain demand.

Fuzzy Logic
Fuzzy Logic is a component of soft computing. Fuzzy logic is an appropriate way to map an input space into an output space. Fuzzy logic uses Fuzzy set theory which considers the degree of membership as a determinant of the state of the elements in a Fuzzy set. Fuzzy is a vague value that can be considered true or false simultaneously. Fuzzy truth and error values depend on the degree of membership possessed by a Fuzzy set. The degree of membership in Fuzzy has a value of zero to one. Fuzzy membership degree with a set that has a zero or one value (yes or no). In a firm set (crisp), the membership value of item x in a set A, which is often written as µA (x), has two possibilities, namely the value of one which means that an item is a member of a set and a zero value which means that an item not become a member in a set. Fuzzy membership has a different probability of having a value at intervals of zero to one but the interpretation of values is very different between the two cases. Fuzzy membership provides a measure of opinion or decision, while probability indicates the proportion of the frequency of an outcome that is true in the long run [3].

Mamdani Method
The Mamdani method is also known as the Min-Max method. Ebrahim Mamdani introduced this method in 1975. Output was obtained through four stages [3].

Formation of Fuzzy Association
Mamdani method input variables and output variables divided into one or more Fuzzy sets, and in each input and output variable there are linguistic variables 2.Application Function Implications Input and output variables are used to determine the function implication. The function implication used in the Mamdani method is to take the minimum value (Min).

Composition of Rules
The composition of the rules obtained from the implication function, which is determining the composition of each rule and the method used in conducting the Fuzzy System Inference, which is the maximum method (Max). In the maximum method (Max) the Fuzzy set solution is obtained by taking the maximum value of the rule, then used to modify the Fuzzy area and applied to the output using the OR (union) operator. If all propositions evaluated, the output will contain a Fuzzy set that reflects the contribution of each proportion.

Defuzzification
Input from the defuzzification process is a set of Fuzzy obtained from the composition of Fuzzy rules, while the output produced is a number of Fuzzy set domains. If a Fuzzy set is given from a certain range, then a certain crisp value is taken as output. In this method crisp solution is obtained by taking the center point of the Fuzzy area. The formula of the Centorid (Composite Moment) method is as follows:

MAE (Mean Absolute Error)
MAE is a model of calculating the error value by calculating the difference between the predicted value and the actual value, which is then authenticated (regardless of positive or negative signs). Results from MAE values in the same form (size) as actual data [4].

MAPE (Mean Absolute Percentage Error)
MAPE is a model of calculating the error value by calculating the difference between the predicted value and the actual value which is then absolute and then calculated in the form of a percentage of the original data (Chang, et al, 2007).

RESULTS AND DISCUSSION
The data taken is data of black parachute fabric raw material that includes data on raw material demand, raw material inventory data and raw material ordering data for three months from January 2019 to March 2019 can be seen in Table 1. Based on Table 1 above, it can be explained that the largest number of fabric raw material requests reached 1971 meters per week and the smallest number of fabric raw material requests reached 306 meters per week. The largest amount of fabric raw material inventory reaches 225 meters per week and the smallest amount of fabric raw material inventory reaches 0 meters per week. The largest number of fabric raw material orders reached 2061 meters per week and the smallest number of fabric raw material orders reached 306 meters per week.

Fuzzy Association Formation
The first step is to determine the variables associated in the process that are determined by the appropriate fuzzification function. This research has input parameters, namely the demand for raw materials and supply of raw materials. Output parameters (orders) are ordered raw materials ordered. In this study there are three variables that are modeled, namely: (8 x 100% = 22 %

DISCUSSION The Result of Ordering of Parachute Fabric Raw Materials
The results of data processing using the Fuzzy Mamdani method obtained prediction of the optimal number of parachute fabric raw materials orders in January the first week of 2019 to March the fourth week of 2019 there are differences in the number of orders between CV. Pine Bag's Specialist and processing results of Fuzzy Mamdani method. The difference in the number of bookings is seen every week for three months.
In January, the first week of ordering parachute fabric raw materials made by CV. The Pine Bag's Specialist is 771 meters smaller than the predicted number of raw materials ordered by the Fuzzy Mamdani method by 879 meters. Number of ordering raw materials CV. Pine Bag's Specialist in January the second week of 1071 meters is greater than the predicted number of Fuzzy Mamdani bookings of 1020 meters. Number of ordering raw materials CV. Pine Bag's Specialist in January the third week of 531 meters was greater than the predicted number of Fuzzy Mamdani method bookings of 438 meters. Number of ordering raw materials CV. Pine Bag's Specialist in January the fourth week of 1881 meters was greater than the predicted number of raw material orders for the Fuzzy Mamdani method by 1440 meters.
In February, the first week of ordering parachute fabric raw materials made by CV. The Pine Bag's Specialist is 621 meters smaller than the predicted number of raw materials ordered by the Fuzzy Mamdani method by 853 meters. Number of ordering raw materials CV. Pine Bag's Specialist in February the second week was 1476 meters smaller than the predicted number of Fuzzy Mamdani bookings of 1750 meters. Number of ordering raw materials CV. Pine Bag's Specialist in February the third week was 733.5 meters smaller than the predicted number of bookings for the Fuzzy Mamdani method by 1000 meters. Number of ordering raw materials CV. Pine Bag's Specialist in February the fourth week was 441 meters smaller than the predicted number of raw material orders for Fuzzy Mamdani method by 561 meters.
In March, the first week of ordering parachute fabric raw materials made by CV. The Pine Bag's Specialist of 2061 meters is greater than the predicted number of raw material orders for the Fuzzy Mamdani method by 1740 meters. Number of ordering raw materials CV. Pine Bag's Specialist in March the second week was 771 meters smaller than the predicted number of Fuzzy Mamdani bookings of 895 meters. Number of ordering raw materials CV. The Pine Bag's Specialist in March the third week was 306 meters smaller than the predicted number of Fuzzy Mamdani method reservations by 418 meters. Number of ordering raw materials CV. Pine Bag's Specialist in March the fourth week was 846 meters smaller than the predicted number of raw material orders for Fuzzy Mamdani method by 1030 meters.

Results of Evaluation Amount of Ordering Parachute Fabric Raw Materials
The results of the evaluation of predictive data on the amount of ordering raw materials Fuzzy Mamdani Method with actual data ordering the number of raw materials for parachute fabric CV. Pine Bag's Specialist which is used to support raw material inventory control decisions in conditions of uncertain demand formed from January the first week of 2019 to March the fourth week of 2019 obtained an average absolute error (MAE) value of 193.8 meters, and the average percentage of absolute error (MAPE) of 22%. According to [5] MAPE is at a value of 20% -50% indicating the level of accuracy of the prediction of the amount of ordering raw materials for parachute fabric with Fuzzy Mamdani said to be reasonable and acceptable.

CONCLUSION
From the results of data processing and discussion conducted, the following conclusions can be drawn: 1. Calculation results of raw material inventory control using the Fuzzy Mamdani method that is done manually, obtained prediction of the number of parachute fabric ordering materials in January the first week of 2019 with a large demand for parachute fabric raw material of 666 meters and 126 meters inventory produces predictions of the number of material orders parachute raw cloth for 879 meters. 2. The evaluation results between the prediction of the amount of ordering parachute fabric raw materials with Fuzzy Mamdani with the actual data ordering fabric raw materials CV. Pine Bag's Specialist from January the first week of 2019 until March 2019 is known to have an average absolute error (MAE) of 193.8 meters with an average percentage of absolute error (MAPE) of 22% so it is considered reasonable to make predictions parachute fabric raw material orders in conditions of uncertain demand in the CV. Pine Bag's Specialist.