An evaluation method for the configuration efficiency of machine-picked cotton processing technology based on slack variables

Abstract

The present invention relates to the field of production process configuration and efficiency evaluation, and specifically to a method for evaluating the efficiency of machine-picked cotton processing process configuration based on slack variables, which includes the following steps: Select the machine-picked cotton processing production line to be evaluated as the decision-making unit, and collect the original index data; Apply the DEA model based on slack variables to calculate the comprehensive efficiency value, technical efficiency value, and scale efficiency value; Determine the non-DEA effective decision-making units according to the comprehensive efficiency value, and respectively judge their technical status and scale returns according to the technical efficiency value and scale efficiency value; Conduct sensitivity analysis on the input and output indicators of the non-DEA effective decision-making units to determine the key process links that need to be improved and determine their improvement amounts. The method provided by the present invention is economical and practical, can evaluate the process configuration efficiency of different machine-picked cotton processing production lines, and further identify the key links that need to be improved, and can provide decision-making references for the implementation of machine-picked cotton processing process design and optimization and improvement.

Description

Technical Field

[0001] The present invention relates to the field of production process configuration and efficiency evaluation, and particularly relates to a method for evaluating the efficiency of machine-picked cotton processing technology based on slack variables. Background Art

[0002] The mechanization of cotton production is one of the key factors for the rapid development of the scale of cotton production in China. Xinjiang is the main cotton-producing area in China. In recent years, the proportion of Xinjiang's cotton output in the total national cotton output has been increasing continuously. In 2020, the proportion of machine-picked cotton output in the total cotton planting output in Xinjiang throughout the year was 81%. The moisture regain rate and impurity content of machine-picked cotton are much higher than those of manually picked cotton. In the process of cotton processing, more seed cotton cleaning, drying, and lint cleaning links are required to produce qualified lint. Therefore, the machine-picked cotton processing link has become increasingly important.

[0003] With the strong promotion of machine-picked cotton planting and processing technologies, it is crucial to improve the supporting machine-picked cotton processing technology for the quality improvement and efficiency increase of the cotton industry. In particular, most of the machine-picked cotton processing lines in China have been developed by transforming hand-picked cotton processing lines or independently researching and developing on the basis of digesting and absorbing foreign technologies. In order to meet the requirements of machine-picked cotton processing, each ginning factory designs or purchases seed cotton processing equipment and supporting equipment according to its own production experience to improve the processing technology. This method of locally optimizing the technology has improved the seed cotton cleaning and processing efficiency and ginning quality to a certain extent, but it has also caused the technology to be complicated and diversified, the configuration of each link to be unreasonable, resulting in high production costs and unstable cotton processing quality.

[0004] The optimization of machine-picked cotton processing technology is a complex systematic project, which requires evaluating the overall efficiency of the processing technology from a systematic perspective, finding the key process links affecting the system efficiency, and carrying out targeted configuration optimization. Therefore, how to quantitatively evaluate the overall efficiency of the machine-picked cotton cleaning and processing technology system, accurately identify the efficiency differences of each process link, and determine the key process procedures to be improved and their improvement amounts are the primary problems to be solved in optimizing the machine-picked cotton processing technology.

[0005] Data Envelopment Analysis (DEA) takes departments or units of the same type (referred to as decision-making units) as the research object and is a method for comprehensively evaluating the effectiveness of processing multi-input and multi-output indicators. This method assumes that each input is associated with one or more outputs, and there is indeed a certain relationship between the inputs and outputs, but it is not necessary to determine the specific expression of this relationship. DEA can evaluate the relative effectiveness among decision-making units with multiple inputs and multiple outputs. Essentially, it judges whether the DMU (decision-making unit) is located on the "production frontier surface" of the production possibility set. It avoids calculating the standard cost of each service because it can convert multiple inputs and multiple outputs into the numerator and denominator of an efficiency ratio without the need to convert them into the same monetary unit. Therefore, using DEA to measure efficiency can clearly illustrate the combination of inputs and outputs. Thus, it is more comprehensive and trustworthy than a set of operating ratios or profit indicators.

[0006] Applying this method to establish a model does not require dimensionless processing of the data, nor does it require subjective setting of weights and parameters. Due to its strong objectivity and convenience in use, it has an absolute advantage in dealing with the effectiveness evaluation of multi-output and multi-input, and has been widely applied.

[0007] In view of the problems existing in the prior art, the present invention starts from a systematic perspective, uses the DEA model to construct an efficiency evaluation system for the machine-picked cotton processing technology, evaluates and analyzes the efficiency of the machine-picked cotton process configuration, finds out the key influencing factors of the machine-picked cotton process configuration efficiency, and is of great significance for adjusting the input resources of the production line and optimizing the process configuration. Summary of the Invention

[0008] The present invention provides an economical and practical method for evaluating the efficiency of the machine-picked cotton processing technology configuration based on slack variables, which can be used to evaluate the overall effectiveness of multiple machine-picked cotton processing production lines from a systematic perspective, and then accurately identify the key process links that need to be improved and their improvement amounts, providing decision-making references for implementing the optimization of the machine-picked cotton processing technology.

[0009] In order to achieve the above object, a method for evaluating the efficiency of the machine-picked cotton processing technology configuration based on slack variables mainly includes the following steps:

[0010] Step 1: Select the machine-picked cotton processing production line to be evaluated as the decision-making unit DMU;

[0011] Step 2: According to the input indicators and output indicators of the machine-picked cotton processing production line, collect the original data of the input indicators and output indicators of each decision-making unit DMU;

[0012] The input indicators include: power input, labor input, seed cotton input, number of seed cotton cleaning times, number of drying times, number of lint cleaning times;

[0013] The output indicators include: the comprehensive quality level of lint cotton and the hourly output level of lint cotton;

[0014] Step 3: Calculate the index value of the comprehensive quality level of lint cotton in the evaluation index system of the machine-picked cotton processing technology configuration efficiency based on the original data;

[0015] The calculation of the index value LQI of the comprehensive quality level of lint cotton is as follows:

[0016] LQI = P + MIC;

[0017] Among them, LQI is the index value of the lint cotton processing quality level, P is the benchmark price of lint cotton in the cotton trading market, and MIC is the contribution of multiple fiber qualities to the lint cotton quality index;

[0018]

[0019] Among them, C i is the contribution of the i-th lint cotton fiber index to the transaction price of lint cotton, and C i can be set according to the "Quality Price Difference Table of Sawtooth Processed Fine Cotton" released by the China Cotton Association;

[0020] Step 4: Input the above input indicators and output indicators into the linear programming solution software for DEA to obtain the comprehensive efficiency value, technical efficiency value, and scale efficiency value of each decision-making unit DMU based on the SE-SBM model;

[0021] Step 5: Determine the non-DEA efficient decision-making units according to the comprehensive efficiency value δ: If the comprehensive efficiency value δ < 1, the i-th machine-picked cotton processing production line is non-DEA efficient; if the comprehensive efficiency value δ ≥ 1, the j-th machine-picked cotton processing production line is DEA efficient;

[0022] Step 6: Judge the technical status and scale returns of the non-DEA efficient decision-making unit DMU according to the pure technical efficiency value and scale efficiency value respectively;

[0023] Technical status judgment: If the technical efficiency value is greater than or equal to 1, it is technically efficient; otherwise, it is inefficient;

[0024] Scale return judgment: If the scale efficiency value is greater than or equal to 1, it is scale efficient; otherwise, it is inefficient;

[0025] Step 7: Conduct a sensitivity analysis on the input and output indicators of the non-DEA efficient decision-making unit, and determine the key process links that need to be improved in combination with the technical status and scale returns;

[0026] (1) Calculate the index sensitivity index S j (X i ):

[0027]

[0028] Among them, A is the original evaluation index set, and A i is the evaluation index set after removing the i-th index, and θ j (A) and θ j (A i ) represent the comprehensive efficiency values of the j-th decision-making unit under the index sets A and A i respectively;

[0029] (2) Judgment of index contribution degree: If the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is max{S j (X i )|i = 1, 2, ……, s}, then the index X i makes the largest DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is relatively reasonable and is well utilized; if the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is min{S j (X i )|i = 1, 2, ……, s}, then the index X i makes the largest non-DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is unreasonable and not fully utilized, and the corresponding process links of the index need to be reasonably adjusted;

[0030] Step 8: Calculate the slack variable values of each input-output index of the non-DEA-effective decision-making unit and determine its improvement amount: Use the linear programming solution software of DEA to calculate the slack variable values of each input-output index of the non-DEA-effective decision-making unit. When the slack variable value is positive, it indicates redundancy, and when it is negative, it indicates deficiency. Determine the improvement amount of the non-DEA-effective decision-making unit based on the slack variable values of the input-output indexes.

[0031] The specific content of step 1, which selects the machine-picked cotton processing production line to be evaluated as the decision-making unit DMU, includes:

[0032] S1: The machine-picked cotton processing production line includes main processes and equipment such as seed cotton cleaning, ginning, lint cleaning, drying and humidifying. The decision-making unit (DMU) regards the machine-picked cotton processing production line as a system composed of multiple inputs and multiple output indexes, and it is called a decision-making unit.

[0033] The specific content of step 2, which collects the original data of each decision-making unit DMU according to the input and output indexes of the machine-picked cotton processing production line, includes:

[0034] S21: Data items of production cost and process configuration input indexes

[0035] (1) Power input: It refers to the power consumption of machine-picked cotton processing equipment operation, cotton processing and transportation, and production auxiliary equipment operation during the processing of machine-picked cotton. The power input index is represented by X1, and the unit is kw·h.

[0036] (2) Labor input: During the processing of machine-picked cotton, personnel are required to supervise equipment parameters, assist in equipment operation, and repair equipment failures. The number of workers selected is used as the labor input index, and the labor input index is represented by X2.

[0037] (3) Seed cotton input: The unit purchase price of seed cotton is affected by indicators such as impurity content, moisture regain, color grade, and length. The purchase price per kilogram of seed cotton is used as the seed cotton input index, and the seed cotton input index is represented by X3, with the unit of yuan / kg.

[0038] (4) Number of seed cotton cleaning times: During the processing of machine-picked cotton, the more the number of seed cotton cleaning processes, the greater the damage to the quality of cotton fibers. The number of seed cotton cleaning times is selected as one of the process configuration indicators, and the seed cotton cleaning times indicator is represented by X4.

[0039] (5) Number of drying times: Drying affects the moisture regain of seed cotton, which in turn affects the ginning quality and energy consumption level of seed cotton. The number of drying times is selected as one of the process configuration indicators, and the number of drying times indicator is represented by X5.

[0040] (6) Number of lint cleaning times: The more the number of lint cleaning processes, the greater the damage to the quality of cotton fibers. The number of lint cleaning times is selected as one of the process configuration indicators, and the number of lint cleaning times indicator is represented by X6.

[0041] S22: Data items of processing quality and lint output indicators

[0042] (1) Comprehensive lint quality level: The processing quality of lint is defined by multiple fiber quality indicators such as color grade, length, micronaire value, breaking ratio strength, length uniformity, ginning quality, and foreign fiber content. The present invention combines the cotton market trading rules and the cotton fiber quality evaluation system to define the comprehensive lint quality index LQI for machine-picked cotton processing, and the comprehensive lint quality level indicator is represented by Y1.

[0043] (2) Lint per-hour output level: The lint per-hour output level indicator reflects the processing efficiency level of the production line. Referring to the standard lint per-hour output range of the gin and the production and processing volume, a production and processing plan is formulated. During the processing, the average hourly processing volume of the gin is the lint per-hour output level, and the lint per-hour output level indicator is represented by Y2.

[0044] The specific steps for calculating the value of the comprehensive lint quality level indicator in the machine-picked cotton processing process configuration efficiency evaluation index system according to the original data in step 3 include:

[0045] S3: Calculate the comprehensive quality index LQI of machine-picked cotton processed lint.

[0046] LQI = P + MIC

[0047] Where, LQI is the lint processing quality index, P is the benchmark price of lint in the cotton trading market, and MIC is the contribution of multiple fiber qualities to the lint quality index.

[0048]

[0049] Where, C i is the contribution of the i-th lint fiber index to the lint transaction price. C i can be set according to the "Quality Price Difference Table for Saw-toothed Processed Fine Cotton" issued by the China Cotton Association.

[0050] Step 4 inputs the input and output indicators into the linear programming solution software for DEA to obtain the comprehensive efficiency value, technical efficiency value, and scale efficiency value of each decision-making unit, specifically including:

[0051] S41: In the machine-picked cotton processing technology configuration efficiency evaluation model SE-SBM (Super Efficiency Slacks-based Measure), regard a machine-picked cotton processing production line and its inputs and outputs as a decision-making unit DMU j . Suppose there are n machine-picked cotton processing technology production lines, and each DMU j (j = 1, 2,..., n) has m inputs and s outputs. X = (x ij ) ∈ R m×n is the input matrix, Y = (y ij ) ∈ R s×n is the output matrix, and X > 0, Y > 0, λ j is the efficiency measurement weight of the decision-making unit.

[0052] S42: Define the production unit set

[0053] Suppose P is the production unit set

[0054]

[0055] S43: Define the production unit subset

[0056] Suppose the production unit subset

[0057]

[0058] S44: Set the weighted average distance

[0059] According to X > 0, Y > 0, we get is a non-empty set. Let L be the weighted average distance from (x0, y0) to and define δ as the value of L.

[0060]

[0061] δ is not less than 1 and equals 1 if and only if when the influence of (x0, y0) on the original production set is excluded. The weighted distance from x0 to represents the average growth rate of x0 to when The weighted distance from y0 to represents the average shrinkage rate of y0 to when

[0062]

[0063]

[0064] where s.t. is the constraint condition, δ is the product of the distance exponents of the input space and the output space, i.e., the comprehensive efficiency value of the decision-making unit, x and y are the input and output items respectively, m and s are the numbers of input and output items respectively, n is the number of decision-making units, and λ is the weight for measuring the efficiency of the decision-making unit.

[0065] The specific steps of determining non-DEA efficient decision-making units according to the comprehensive efficiency value in step 5 include:

[0066] S5: For a non-DEA efficient decision-making unit (DMU j ), it is determined that

[0067] If δ < 1, then the jth machine-picked cotton processing production line (DMU j ) is non-DEA efficient, indicating that the input-output ratio of the production line is unreasonable and the process configuration needs to be adjusted; if δ ≥ 1, then the jth machine-picked cotton processing production line (DMU j ) is DEA efficient, indicating that the production input is reasonable.

[0068] The specific steps of judging the technical status and scale returns of non-DEA efficient decision-making unit DMU according to the pure technical efficiency value and the scale efficiency value in step 6 include:

[0069] S61: Pure technical efficiency judgment: If the technical efficiency value is greater than or equal to 1, it is technically efficient; otherwise, it is inefficient. Conduct a specific analysis of the production line with relatively low pure technical efficiency and carry out further upgrading and transformation.

[0070] S62: Scale efficiency judgment: If the scale efficiency value is greater than or equal to 1, it is scale efficient; otherwise, it is inefficient.

[0071] ​Step 7 performs sensitivity analysis on the input and output indicators of non-DEA effective decision-making units, and determines the specific key process links that need to be improved by combining the technical status and scale returns, specifically including:

[0072] S71: Calculate the index sensitivity index S j (X i ):

[0073]

[0074] where A is the original evaluation index set, and A i is the evaluation index set after removing the i-th index, and θ j (A) and θ j (A i ) represent the comprehensive efficiency values of the j-th decision-making unit under the index sets A and A i ;

[0075] S72: Index contribution degree judgment: If the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is max{S j (X i )|i = 1, 2, ……, s}, then the index X i has the greatest DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is relatively reasonable and is well utilized; if the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is min{S j (X i )|i = 1, 2, ……, s}, then the index X i has the greatest non-DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is unreasonable and not fully utilized, and the corresponding process link of the index needs to be reasonably adjusted;

[0076] S8: Calculate the slack variable values of each input and output indicator of non-DEA effective decision-making units, and determine their improvement amounts, specifically including: Using the linear programming solution software of DEA to calculate the slack variable values of non-DEA effective process input and output indicators. Positive values indicate redundancy, and negative values indicate deficiency. Determine the improvement amount of non-DEA effective processes based on the slack variable values of input and output indicators.

[0077] The present invention has the following advantages:

[0078] The method provided by the present invention is economical and practical, can evaluate the process configuration efficiency of different machine-picked cotton processing production lines, and then identify the key links that need to be improved, and can provide decision-making references for the implementation of machine-picked cotton processing process design and optimization improvement.

[0079] (1) Based on the current situation of machine - picked cotton processing technology and the influencing factors of process efficiency, the present invention determines the evaluation index of machine - picked cotton processing technology efficiency and constructs an evaluation index system for the configuration efficiency of machine - picked cotton processing technology;

[0080] (2) Aiming at quantitatively evaluating the configuration efficiency of machine - picked cotton processing technology, the index system and evaluation method constructed by using the data envelopment analysis method can be used to evaluate the configuration efficiency of different machine - picked cotton processing production lines, analyze the improvement amount of non - DEA - effective processes, find out the differences in the configuration efficiency of each processing link, and provide decision - making reference for the implementation of machine - picked cotton processing technology design and improvement.

[0081] From a systematic perspective, the present invention uses the DEA model to construct an evaluation system for machine - picked cotton processing technology efficiency, evaluates and analyzes the configuration efficiency of machine - picked cotton technology, finds out the key influencing factors of machine - picked cotton technology configuration efficiency, and is of great significance for adjusting the input resources of the production line and optimizing the process configuration. Brief Description of the Drawings

[0082] The present invention will be further described below in conjunction with the drawings and examples.

[0083] Figure 1 is the flow block diagram of the present invention.

[0084] Figure 2 is the machine - picked cotton processing technology configuration efficiency diagram of the example of the present invention.

[0085] Figure 3 is the machine - picked cotton processing technology index sensitivity index diagram of the example of the present invention.

[0086] Figure 4 is the machine - picked cotton non - DEA - effective processing technology output slack variable diagram of the example of the present invention. Detailed Embodiment

[0087] The technical solution of the present invention will be further specifically described below through examples and in conjunction with the drawings:

[0088] Example:

[0089] A method for evaluating the configuration efficiency of machine - picked cotton processing technology based on slack variables mainly includes the following steps:

[0090] Step 1: Select the machine - picked cotton processing production line to be evaluated as the decision - making unit DMU;

[0091] Step 2: According to the input indexes and output indexes of the machine - picked cotton processing production line, collect the original data of the input indexes and output indexes of each decision - making unit DMU;

[0092] The input indicators include: power input, labor input, seed cotton input, number of seed cotton cleaning times, number of drying times, and number of lint cleaning times;

[0093] The output indicators include: comprehensive lint quality level and lint hourly output level;

[0094] Step 3: Calculate the index value of the comprehensive lint quality level in the evaluation index system of the machine-picked cotton processing technology configuration efficiency based on the original data;

[0095] The calculation of the index value LQI of the comprehensive lint quality level is as follows:

[0096] LQI = P + MIC;

[0097] Among them, LQI is the index value of the lint processing quality level, P is the benchmark price of lint in the cotton trading market, and MIC is the contribution of multiple fiber qualities to the lint quality index;

[0098]

[0099] Among them, C i is the contribution of the i-th lint fiber index to the lint transaction price, and C i can be set according to the "Quality Price Difference Table of Saw-toothed Processed Fine Cotton" issued by the China Cotton Association;

[0100] Step 4: Input the above input indicators and output indicators into the linear programming solution software for DEA to obtain the comprehensive efficiency value, technical efficiency value, and scale efficiency value of each decision-making unit DMU based on the SE-SBM model;

[0101] Step 5: Determine the non-DEA effective decision-making units according to the comprehensive efficiency value δ: If the comprehensive efficiency value δ < 1, the i-th machine-picked cotton processing production line is non-DEA effective; if the comprehensive efficiency value δ ≥ 1, the j-th machine-picked cotton processing production line is DEA effective;

[0102] Step 6: Judge the technical status and scale return of the non-DEA effective decision-making unit DMU according to the pure technical efficiency value and scale efficiency value respectively;

[0103] Technical status judgment: If the technical efficiency value is greater than or equal to 1, it is technically effective; otherwise, it is ineffective;

[0104] Scale return judgment: If the scale efficiency value is greater than or equal to 1, it is scale effective; otherwise, it is ineffective;

[0105] Step 7: Conduct a sensitivity analysis on the input and output indicators of the non-DEA effective decision-making unit, and determine the key process links that need to be improved in combination with the technical status and scale return;

[0106] (1) Calculate the index sensitivity index Sj (X i ):

[0107]

[0108] Among them, A is the original evaluation index set, and A i is the evaluation index set after removing the i-th index, and θ j (A) and θ j (A i ) represent the comprehensive efficiency values of the j-th decision-making unit under the index sets A and A i respectively;

[0109] (2) Judgment of index contribution degree: If the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is max{S j (X i )|i = 1, 2,..., s}, then the index X i makes the largest DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is relatively reasonable and is well utilized; if the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is min{S j (X i )|i = 1, 2,..., s}, then the index X i makes the largest non-DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is unreasonable and not fully utilized, and the corresponding process links of the index need to be reasonably adjusted;

[0110] Step 8: Calculate the slack variable values of each input-output index of the non-DEA-effective decision-making unit and determine its improvement amount: Use the linear programming solution software of DEA to calculate the slack variable values of each input-output index of the non-DEA-effective decision-making unit. When the slack variable value is positive, it indicates redundancy, and when it is negative, it indicates deficiency. Determine the improvement amount of the non-DEA-effective decision-making unit based on the slack variable values of the input-output indexes.

[0111] The specific steps of step 1 include:

[0112] S1: Regard the machine-picked cotton processing production line as a system composed of multiple inputs and multiple output indexes, and call it the decision-making unit DMU.

[0113] The specific steps of step 2 include:

[0114] S21: Data items of production cost and process configuration input indexes

[0115] (1) Power input; The power input index is represented by X1, and the unit is kw·h.

[0116] (2) Labor input; the number of workers is selected as the labor input index, and the labor input index is represented by X2.

[0117] (3) Seed cotton input; the purchase price of seed cotton is used as the seed cotton input index, represented by X3, in yuan / kg.

[0118] (4) Number of times of seed cotton cleaning; the index of the number of times of seed cotton cleaning is represented by X4.

[0119] (5) Number of drying times; the index of the number of drying times is represented by X5.

[0120] (6) Number of times of lint cleaning; the index of the number of times of lint cleaning is represented by X6.

[0121] S22: Data items of output indicators for processing quality and lint yield

[0122] (1) Comprehensive quality level of lint

[0123] The comprehensive quality index of lint processed from machine-picked cotton is LQI, and the comprehensive quality level index of lint is represented by Y1.

[0124] (2) Lint output per machine-hour level; the average amount of lint processed per hour by the gin during the processing is the lint output per machine-hour level, and the lint output per machine-hour level index is represented by Y2.

[0125] The specific steps of step 3 include:

[0126] S3: Calculation of the comprehensive quality index LQI of lint processed from machine-picked cotton.

[0127] LQI = P + MIC

[0128] Among them, LQI is the lint processing quality index, P is the benchmark price of lint in the cotton trading market, and MIC is the contribution of multiple fiber qualities to the lint quality index.

[0129]

[0130] Among them, C i is the contribution of the i-th lint fiber index to the lint transaction price. C i can be set according to the "Quality Price Difference Table of Saw-toothed Processed Fine Cotton" issued by the China Cotton Association.

[0131] The specific steps of step 4 include:

[0132] S4: Define the set of production units and subsets of production units, then set the weighted average distance to obtain the comprehensive efficiency value, pure technical efficiency value, and scale efficiency value of the decision-making unit DMU under the SE-SBM model. The calculation results are as Figure 2As shown in the figure, the processing technology representation method is: number of seed cleaning times - number of drying times - number of lint cleaning times, ginning machine model. For example, the processing technology code of decision-making unit A is 2-1-1MY96, indicating that the processing technology of decision-making unit A is 2 times of seed cleaning, 1 time of drying, 1 time of lint cleaning, and the ginning machine model is MY96.

[0133] Step 5 specifically includes:

[0134] S5: If δ < 1, then the jth machine-picked cotton processing production line (DMU j ) is non-DEA efficient; if δ ≥ 1, then the jth machine-picked cotton processing production line (DMU j ) is DEA efficient. As Figure 3 can be seen, the process configuration efficiency of different production lines is sorted as C > A > E > G > H > I > F > D > B according to the comprehensive efficiency value. Among them, the process efficiency values of production lines H, I, F, D, and B are less than 1, which are DEA inefficient, and the process efficiency values of production lines C, A, E, and G are greater than 1, which are DEA efficient. This indicates that the inputs of production lines C, A, E, and G are reasonable, while the input-output ratios of production lines H, I, F, D, and B are unreasonable, and the process configuration needs to be adjusted.

[0135] Step 6 specifically includes:

[0136] S61: Judgment of pure technical efficiency. As Figure 3 can be seen, the overall pure technical efficiency value of the machine-picked cotton processing technology is relatively high, with an average value of 0.9693, indicating that certain achievements have been made in the process technology transformation of cotton processing enterprises. However, there are significant differences in the technical levels and management effects of different production lines. Among them, the pure technical efficiencies of production lines B and F are both less than 0.8, and the pure technical efficiency is relatively low. Specific analysis needs to be carried out on production lines B and F to find the reasons for the backward technical equipment and carry out further upgrading and transformation.

[0137] S62: Analysis of scale efficiency. As Figure 3 can be seen, the average scale efficiency of each machine-picked cotton processing production line is 1.014, and the efficiency value is stable, indicating that the overall scale of most cotton processing production lines has reached a relatively reasonable level. From the perspective of scale returns, production lines A, C, E, G, and H have constant returns to scale, and these five production lines are in the best state of scale returns; production lines B, F, and I have increasing returns to scale. When the input indicators increase, not only the output increases, but also the "speed" of increase is increasing, and the production line scale can be expanded; production line D has decreasing returns to scale. When the input indicators increase, the output increases, but the "speed" of increase is decreasing, and the scale of production line D should not be increased.

[0138] Step 7 specifically includes:

[0139] S71: Calculate the index sensitivity index S j (Xi ):

[0140]

[0141] Among them, A is the original evaluation index set, and A i is the evaluation index set after removing the i-th index, and θ j (A) and θ j (A i ) represent the comprehensive efficiency of the j-th decision-making unit under the index sets A and A i .

[0142] S72: Index contribution degree judgment: If the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line (DMU j ) is max{S j (X i )|i = 1, 2, ……, s}, then the DEA effective contribution of the index X i to the j-th machine-picked cotton processing production line (DMU j ) is the largest; if the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line (DMU j ) is min{S j (X i )|i = 1, 2, ……, s}, then the non-DEA effective contribution of the index X i to the j-th machine-picked cotton processing production line (DMU j ) is the largest.

[0143] S73: Determine the key improvement processes according to the index contribution degree judgment result, technical status and scale return.

[0144] As Figure 3 shown, in terms of input indicators, ∑S j (X2) = 1.1863 is the largest, and ∑S j (X4) = 0.2511 is the smallest, indicating that the seed cotton input index has the largest effective contribution to the decision-making unit, and the seed cotton is well utilized. The seed cotton cleaning times index has the largest relative ineffective contribution, and the seed cotton cleaning times is unreasonable. In terms of output indicators, ∑S j (Y1 = 1.9230) is the largest, and ∑S j (Y2) = 0.5275 is the smallest, indicating that the lint processing quality is relatively reasonable and the production input resources are fully utilized. The lint output per machine-hour level is unreasonable and the production line process configuration is not fully utilized. Therefore, to improve the process efficiency of the production line, it is necessary to reasonably adjust the seed cotton cleaning process number index and the lint output per machine-hour level index.

[0145] The specific steps of step 8 are as follows:

[0146] S8: Use MaxDEA Ultra software to calculate the slack variables of the input-output indicators of non-DEA-efficient processes. Analyze the improvement amount of non-DEA-efficient processes based on the slack variables of the input-output indicators. The calculation results are as Figure 4 shown.

[0147] From Figure 4 the analysis, it can be seen that there are redundancies and deficiencies in some indicators of the input indicator system and output indicator system of non-DEA-efficient processes B, D, F, H, and I. To upgrade the DEA-inefficient processes to DEA-efficient, B, D, F, H, and I need to make the following adjustments in terms of input and output:

[0148] (1) In terms of power input, there are redundancies in B, D, F, H, and I, and the electricity consumption can be reduced by 26%, 53%, 50%, 50%, and 21% respectively; in terms of labor input, there are redundancies in D, F, H, and I, and the number of workers can be reduced by 12.5%, 5.6%, 6.3%, and 16% respectively.

[0149] (2) In terms of process configuration, there are redundancies in the production lines of B, D, and I, and the number of times of seed cotton cleaning can be reduced by 1 time, 2 times, and 1 time respectively; there are redundancies in the production lines of B and F, and the number of times of lint cleaning and drying can be reduced by 1 time each. After improvement, the five production lines are all configured with a 3-1-1 process.

[0150] (3) In terms of the lint hourly output level, there are deficiencies in the production lines of B, D, F, H, and I. Under the condition of ensuring the processing quality of machine-picked cotton, there is a potential for increasing the lint hourly output by 73%, 19%, 32%, 36%, and 13% respectively. From the perspective of process configuration, there is overcapacity or insufficient processing volume of machine-picked cotton in the 5 production lines, and the production capacity should be reduced or the processing volume of seed cotton should be increased.

Claims

1. A method for evaluating the efficiency of the processing technology configuration of machine-picked cotton based on slack variables, characterized in that, It includes the following steps: Step 1: Select the machine-picked cotton processing production line to be evaluated as the decision-making unit DMU; Step 2: According to the input indicators and output indicators of the machine-picked cotton processing production line, collect the original data of the input indicators and output indicators of each decision-making unit DMU; The input indicators include: electricity input, labor input, seed cotton input, number of seed cotton cleaning times, number of drying times, number of lint cleaning times; The output indicators include: comprehensive lint quality level, lint hourly output level; Step 3: Calculate the index value of the comprehensive lint quality level in the evaluation index system of the machine-picked cotton processing technology configuration efficiency according to the original data; The calculation of the index value LQI of the comprehensive lint quality level is as follows: LQI = P + MIC; Among them, LQI is the index value of the lint processing quality level, P is the benchmark price of lint in the cotton trading market, and MIC is the contribution of multiple fiber qualities to the lint quality index; Among them, C i is the contribution of the i-th lint fiber index to the transaction price of lint, and C i is set according to the "Quality Price Difference Table of Saw-ginned Fine Cotton" released by the China Cotton Association; Step 4: Input the above input indicators and output indicators into the linear programming solution software for DEA to obtain the comprehensive efficiency value, technical efficiency value and scale efficiency value of each decision-making unit DMU based on the SE-SBM model; Step 5: Determine the non-DEA efficient decision-making unit according to the comprehensive efficiency value δ: If the comprehensive efficiency value δ < 1, the i-th machine-picked cotton processing production line is non-DEA efficient; if the comprehensive efficiency value δ ≥ 1, the j-th machine-picked cotton processing production line is DEA efficient; Step 6: Judge the technical status and scale return of the non-DEA efficient decision-making unit DMU according to the pure technical efficiency value and scale efficiency value respectively; Technical status judgment: If the technical efficiency value is greater than or equal to 1, it is technically efficient; otherwise, it is inefficient; Scale return judgment: If the scale efficiency value is greater than or equal to 1, it is scale efficient; otherwise, it is inefficient; Step 7: Conduct a sensitivity analysis on the input and output indicators of the non-DEA efficient decision-making unit, and determine the key process links that need to be improved in combination with the technical status and scale return; (1) Calculate the index sensitivity index S j (X i ): Among them, A is the original evaluation index set, A i is the evaluation index set after removing the i-th index, θ j (A) and θ j (A i ) represents the comprehensive efficiency value of the j-th decision-making unit under the index sets A and A i respectively; (2) Index contribution degree judgment: If the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is max{S j (X i )|i = 1, 2, ……, s}, then the index X i has the largest DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is relatively reasonable and is well utilized; if the sensitivity index of the i-th index X i to the j-th machine-picked cotton processing production line is min{S j (X i )|i = 1, 2, ……, s}, then the index X i has the largest non-DEA-effective contribution to the j-th machine-picked cotton processing production line, indicating that the index X i is unreasonable and not fully utilized, and the corresponding process links of the index need to be reasonably adjusted; Step 8: Calculate the slack variable values of each input and output indicator of the non-DEA efficient decision-making unit to determine its improvement amount: Use the linear programming solution software of DEA to calculate the slack variable values of each input and output indicator of the non-DEA efficient decision-making unit. When the slack variable value is positive, it indicates redundancy, and when it is negative, it indicates deficiency. Determine its improvement amount according to the slack variable values of the input and output indicators of the non-DEA efficient decision-making unit.

2. The method for evaluating the configuration efficiency of the machine-picked cotton processing technology based on slack variables according to claim 1, wherein, The machine-picked cotton processing production line includes process equipment such as seed cotton cleaning, ginning, lint cleaning, drying, and humidifying. The decision-making unit DMU regards the machine-picked cotton processing production line as a system composed of multiple input indicators and multiple output indicators, which is called the decision-making unit, abbreviated as DMU.

Images