Self-adaptive control method of full-automatic washing packing scale unit

Abstract

The invention discloses a self-adaptive control method for a full-automatic washing packing scale unit, and relates to the technical field of intelligent control in the processing and production industry. The actions of all parts of the full-automatic washing unit and the quantitative packaging part are coordinated and controlled by a PID (Proportion Integration Differentiation) controller based on an improved saltcat optimization algorithm; in an improved Karat optimization algorithm, a chaotic mapping strategy is introduced to initialize a Karat population, an adaptive convergence factor is introduced in an exploration stage to improve the search efficiency, a level flight strategy is introduced to improve the global search capability, and a sparrow alert mechanism is added to improve the convergence speed. According to the washing and weighing integrated packing scale, integration of washing, weighing and packing functions is achieved for the first time; and a control system of the full-automatic washing packing scale is designed, so that intelligent material out-of-tolerance correction and excessive and insufficient alarm are realized, and the production efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent control in processing and production industries, and in particular relates to a self-adaptive control method for a full-automatic water-washing packaging scale unit. Background Art

[0002] Packaging scales are an important equipment widely used in the field of industrial production. They are mainly used to automatically complete weighing, metering, packaging and other processes during the commodity packaging process. With the popularization of large-scale and intelligent production, fully automatic packaging scales play a key role in improving production efficiency, reducing labor costs, and ensuring product quality. They have high demand and broad market prospects.

[0003] Although fully automatic scales have improved in terms of automation and accuracy, the following problems still exist: (1) Existing fully automatic scales lack adaptability during the weighing process and cannot be flexibly adjusted according to the characteristics of different materials and packaging conditions. This often requires manual setting of a series of parameters when processing different commodities, increasing the complexity and time cost of the operation; (2) Since packaging scales are affected by factors such as the environment, product shape, and packaging materials in actual use, there may be certain errors in the weighing results; existing fully automatic scales lack real-time monitoring and correction mechanisms for weighing errors and cannot guarantee the accuracy and stability of the weighing results; (3) Traditional fully automatic scales have certain difficulties in the cleaning process, which requires a lot of time and manpower; this limits their continuous use and efficiency on the production line; therefore, this technical solution proposes an adaptive control method for a fully automatic water-washing packaging scale unit. Summary of the invention

[0004] The invention provides an adaptive control method for a full-automatic water-washing packaging scale unit, which solves the problems in the background technology.

[0005] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The adaptive control method of a fully automatic water-washing packaging scale unit of the present invention is implemented based on the fully automatic water-washing packaging scale unit of ISCSO and its adaptive control system. The structure of the fully automatic water-washing packaging scale unit includes: a water-washing packaging scale shell, a storage hopper installed on the upper part of the water-washing packaging scale shell, a fully automatic water-washing unit, a load-bearing unit and a quantitative packaging unit; the fully automatic water-washing unit is integrated in the water-washing packaging scale shell, and the fully automatic water-washing unit includes a water / air washing pipeline and an arc-shaped guide cleaning plate module; wherein the storage hopper is used to receive materials to ensure that the materials fall into the packaging bag, and detect the weight of the materials in the bag at the load-bearing unit, transmit the data to the control system, fine-tune the material quality, and perform appropriate packaging in the quantitative packaging part. At the same time, the fully automatic water-washing unit cleans the inside of the packaging scale to ensure the accuracy of the weighing.

[0007] The adaptive control method of the fully automatic water-washing packaging scale unit comprises the following steps:

[0008] Step 1: Analyze the unloading module of the fully automatic water-washing packaging scale, study the relationship between feeding and weight, and establish a fully automatic water-washing packaging controller;

[0009] Step 2: Design an improved sand cat optimization algorithm and introduce three improvement points. In the population initialization stage, the Piecewise chaotic mapping theory is introduced to initialize the sand cat population. In the attack stage, the Levy flying and wandering strategy is introduced. Since the distance between the sand cat and the prey is very close when attacking the prey, in order to be more in line with the behavior of the sand cat attacking the prey, the adjustment coefficient C=0.37 is introduced to make the sand cat wander as close to the prey as possible.

[0010] Step 3: After the hunting phase of the sand cats, a sparrow alert mechanism is introduced. When the sand cat group realizes the danger, the edge sand cats quickly move to the safe area, and the sand cats in the population randomly move close to other sand cats, thereby improving the convergence speed of the sand cat optimization algorithm.

[0011] Step 1 is specifically:

[0012] The mathematical model of the packaging scale is established according to the principles of mechanics. Since it is impossible to accurately detect the impact force I(t) generated by the falling material at any time, it is difficult to accurately predict the shutdown time t, so an online adjustment strategy is adopted.

[0013] Assume that at the shutdown time t, the measured weight of the nth bag of material is M(n), the mass of the falling material is Ml(n), the impact force generated by the falling material is I(n), the actual weight of the material is S(n), the actual deviation at this time is e(n), and the nominal quantitative value is S, then:

[0014] S(n)=M(n)+Ml(n)-I(n) (1)

[0015] e(n)=SS(n) (2)

[0016] e(n)=SM(n)+[I(n)-Ml(n)] (3)

[0017] Let X(n) = [I(n) - Ml(n)], where X(n) represents the algebraic sum of the impact force of the falling material and the weight of the falling material in the air when weighing the nth bag of material, that is:

[0018] e(n)=SM(n)+X(n), n=1,2,3,...,e(0)=0 (4)

[0019] Obviously,

[0020] M(n)=M0 +e(n - 1) / 2 (5)

[0021] Where M 0 is a fixed value, and 2 / 3S < M 0 < S, there is

[0022] e(n) = S - M 0 -e(n - 1) / 2 + X(n) (6)

[0023] e(n - 1) = S - M 0 -e(n - 2) / 2 + X(n - 1) (7)

[0024] Subtracting equation (7) from equation (6) gives the iterative formula:

[0025] e(n) - e(n - 1) = 1 / 2[e(n - 1) - e(n - 2)] + [X(n) - X(n - 1)] (8)

[0026] e(n) - e(1) = 1 / 2[e(n - 1) - e(0)] + [X(n) - X(1)] = -1 / 2e(n - 1) + [X(n) - X(1)] (9)

[0027] e(n) = X(n) - 1 / 2e(n - 1) + e(1) - X(1) (10)

[0028] From equation (1), e(1) = S - M(1) + X(1), then:

[0029] e(n) = -1 / 2e(n - 1) + [X(n) + S - M(1)] (11)

[0030] From equation (2), M(1) = M 0 , then:

[0031] e(n) = -1 / 2e(n - 1) + [X(n) + S - M 0 (12)

[0032] Through the above derivation, the current weight deviation of the material is obtained, the mathematical model of the fully automatic water - washing packaging scale is established, and the control is carried out in the way of fuzzy PID;

[0033] Specifically, step two is:

[0034] The improved sand cat optimization algorithm is used to optimize and select the control parameters of PID. Among them, the basic sand cat optimization algorithm simulates the hunting behavior of sand cats, mainly including two aspects: searching and predation. Its algorithm flow is as follows:

[0035] First, initialize the population:

[0036] X i = l b +a×(u b -l b ) (13)

[0037] Where: X i represents the position of the i-th sand cat individual, a is a random number in [0,1], u b represents the upper bound, l b Indicates the lower bound.

[0038] Prey Search Phase:

[0039] The sand cat searches for prey by emitting low-frequency noise signals. The signal sensing range is from 2kHz to 0. As the iteration proceeds, the signal value r G It decreases linearly from 2 to 0, simulating the process of a sand cat gradually approaching its prey without losing or jumping. M Imitates the auditory characteristics of the sand cat:

[0040]

[0041] t is the current iteration number, and T is the maximum iteration number. This value can be flexibly set according to different problems to be solved, which shows the universality of the proposed equation.

[0042] The main parameter controlling the transition between exploration and development process is R:

[0043] R = 2 × r G ×rand() (15)

[0044] Among them, rand() represents a random number between 0 and 1, r G Represents the process range where the sensitivity decreases linearly from 2 to 0.

[0045] In the search step, the update of each sand cat's current search position is based on a random position. This method allows the sand cat to explore new spaces in the search space. In order to prevent falling into the local optimum, each sand cat has a different sensitivity range r 2 :

[0046] r 2 =r G ×rand() (16)

[0047] Therefore, r 2 Used for position updates during exploration and development phases, r G It is used to guide the parameter R to switch between two stages.

[0048] In the sand cat optimization algorithm, the sand cat will update its position according to the optimal position of the current population, its own position and the sensitivity range. In order to ensure that the sand cat finds a new optimal position in the search area, which is between the current position and the target prey position, and at the same time ensure the convergence effect and complexity of the algorithm, the following position update mathematical model is established:

[0049]

[0050] where x t best Represents the position of the best individual in the current sand cat population.

[0051] Predation stage:

[0052] The sand cat identifies prey based on the signals it hears. The position update model at this stage is as follows:

[0053]

[0054] Where rand() represents a random number between [0,1]. The search range of the sand cat is considered to be a circle. Its moving direction can be defined by a random angle θ on the circle, ranging from 0 to 360 degrees. Each sand cat in the population can move along a different circular direction to avoid falling into a local optimum. P is the distance the sand cat moves in one step, and x i t and x i t+1 are the positions of the i-th sand cat at time t and time t+1 respectively.

[0055] Exploration and development of sand cat populations:

[0056] Exploration and development of sand cat colonies by R and r G The adaptive value of ensures that the Sand Cat Algorithm performs development (hunting) when the R value is less than or equal to 1, otherwise it can only perform exploration (search):

[0057]

[0058] Formula (19) represents the position update method of the i-th sand cat in the exploration and development stages. The different search methods of each sand cat prevent the algorithm from falling into the local optimum and ensure the global search capability.

[0059] Piecewise chaotic mapping is introduced to optimize the initialization of sand cat population:

[0060]

[0061] Introducing the Levy flight and wandering strategy during the hunting process of the sand cat can provide a wandering method with a step length that conforms to the Levy distribution. However, during the predation stage, the distance between the sand cat and the prey is very close. In order to be more consistent with the predation behavior of the sand cat, the adjustment constant C = 0.37 is introduced to make the sand cat wander as close to the prey as possible. The Levy flight strategy formula is as follows:

[0062]

[0063] Step three is as follows:

[0064] After the sand cat predation phase, a sparrow alert mechanism is introduced to prevent the sand cat from being discovered by its predators, thereby improving the convergence speed of the sand cat optimization algorithm.

[0065] When sand cats are aware of danger, those at the edge of the group will quickly move to a safe area to gain a better position, while those in the middle of the group will move randomly and get closer to other sand cats:

[0066]

[0067] Among them, x t best and x t worst Represent the current global best position and worst position respectively. β is a random number with controllable step size that obeys the normal distribution [0,1]. k represents the direction of the sand cat's movement and is a uniform random number in the range [-1,1]. and represent the current global best and worst fitness values ​​respectively, and ε is the minimum constant to avoid the denominator being 0. When f i =f g This means that the sand cat is on the edge of the population and is vulnerable to attacks by predators. It should be moved to a safe place as soon as possible.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] (1) This solution aims to solve the defects of existing fully automatic packaging scales, such as material residue affecting accuracy and inconvenient cleaning. We innovatively proposed a fully automatic water washing unit, which can quickly and residue-free by dual cleaning with water and air. We also conducted a mechanical analysis of feeding of the fully automatic water washing packaging scale, studied the effect of feeding time on material weight, and established an adaptive controller for the water washing packaging scale.

[0070] (2) This scheme designs an improved sand cat optimization algorithm ISCSO, which introduces three major improvements: In the population initialization stage, the Piecewise chaotic mapping strategy is introduced to initialize the sand cat population; In the population predation stage, the Levy flight strategy is introduced to improve the global optimization ability of the algorithm, and the adjustment constant is introduced to ensure the search speed; After the predation stage, the sparrow alert strategy is introduced to make the sand cats in the middle of the population move closer to the optimal position, and the sand cats in the worst position stay away from the edge of the population, ensuring the search efficiency;

[0071] (3) Through the above optimization, the performance of the Sand Cat algorithm is effectively improved, avoiding falling into the local optimum, and achieving the control accuracy and adaptive adjustment ability of the fully automatic water-washing packaging scale unit. The experiment of this method provides a theoretical reference for the precise control of the weight of production packaging and other links, which has certain practical significance.

[0072] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0074] Figure 1 It is a structural schematic diagram of a fully automatic water-washing packaging scale unit of the present invention;

[0075] Figure 2 An adaptive control system diagram corresponding to an adaptive control method of a fully automatic water-washing packaging scale unit of the present invention;

[0076] Figure 3 A flowchart of a self-adaptive control method for a fully automatic water-washing packaging scale unit according to the present invention;

[0077] Figure 4 It is a convergence curve diagram of the first test function in the embodiment of the present invention under the improved sand cat optimization algorithm and other multiple intelligent optimization algorithms;

[0078] Figure 5 It is a convergence curve diagram of the second test function in the embodiment of the present invention under the improved sand cat optimization algorithm and other multiple intelligent optimization algorithms;

[0079] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0080] 1. Storage hopper; 2. Water-washing packaging scale housing; 3. Load-bearing unit; 4. Quantitative packaging unit; 5. Control unit; 6. Water / air washing pipeline. DETAILED DESCRIPTION

[0081] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0082] like Figure 1 As shown, an adaptive control method of a fully automatic water-washing packaging scale unit of the present invention is implemented based on ISCSO. The fully automatic water-washing packaging scale unit includes a storage hopper 1, a fully automatic water-washing unit, a load-bearing unit 3, and a quantitative packaging unit 4. The fully automatic water-washing unit is integrated into the packaging scale for the first time, and includes a water and air intelligent spraying module 6 and an arc-shaped guide cleaning plate module;

[0083] The working control of the fully automatic water-washing packaging scale includes the following steps:

[0084] Step 1: Analyze the unloading module of the fully automatic water-washing packaging scale, study the relationship between feeding and weight, and establish a fully automatic water-washing packaging controller;

[0085] Step 2: Design an improved sand cat optimization algorithm and introduce three improvement points. In the population initialization stage, the Piecewise chaotic mapping theory is introduced to initialize the sand cat population. In the attack stage, the Levy flying and wandering strategy is introduced. Since the distance between the sand cat and the prey is very close when attacking the prey, in order to be more in line with the behavior of the sand cat attacking the prey, the adjustment coefficient C = 0.37 is introduced to make the sand cat wander around the prey as much as possible; the adjustment coefficient is less than 1 to reduce the search time, and this value is an empirical value.

[0086] Step 3: After the sand cat hunting phase is over, a sparrow alert mechanism is introduced. When the sand cat group realizes the danger, the edge sand cats quickly move to the safe area, and the sand cats in the population randomly move close to other sand cats, thereby improving the convergence speed of the sand cat optimization algorithm.

[0087] Step 1 is specifically:

[0088] The mathematical model of the packaging scale is established according to the principles of mechanics. Since it is impossible to accurately detect the impact force I(t) generated by the falling material at any time, it is difficult to accurately predict the shutdown time t, so an online adjustment strategy is adopted.

[0089] Assume that at the shutdown time t, the measured weight of the nth bag of material is M(n), the mass of the falling material is Ml(n), the impact force generated by the falling material is I(n), the actual weight of the material is S(n), the actual deviation at this time is e(n), and the nominal quantitative value is S, then:

[0090] S(n) = M(n) + Ml(n) - I(n) (1)

[0091] e(n) = S - S(n) (2)

[0092] e(n) = S - M(n) + [I(n) - Ml(n)] (3)

[0093] Let X(n) = [I(n) - Ml(n)], X(n) represents the algebraic sum of the impact force of the falling material and the weight of the falling material in the air when weighing the nth bag of material, that is:

[0094] e(n) = S - M(n) + X(n), n = 1, 2, 3,..., e(0) = 0(

[0095] Obviously, it can be known that:

[0096] M(n) = M 0 + e(n - 1) / 2(5

[0097] where M 0 is a fixed value, and 2 / 3S < M 0 < S, there is

[0098] e(n) = S - M 0 - e(n - 1) / 2 + X(n) (6)

[0099] e(n - 1) = S - M 0 - e(n - 2) / 2 + X(n - 1) (7)

[0100] Subtracting equation (7) from equation (6) gives the iterative formula:

[0101] e(n) - e(n - 1) = 1 / 2[e(n - 1) - e(n - 2)] + [X(n) - X(n - 1)] (8)

[0102] e(n) - e(1) = 1 / 2[e(n - 1) - e(0)] + [X(n) - X(1)] = -1 / 2e(n - 1) + [X(n) - X(1)] (9)

[0103] e(n) = X(n) - 1 / 2e(n - 1) + e(1) - X(1) (10)

[0104] From equation (1), e(1) = S - M(1) + X(1), then:

[0105] e(n) = -1 / 2e(n - 1) + [X(n) + S - M(1)] (11)

[0106] From equation (2), M(1) = M0 ,but:

[0107] e(n)=-1 / 2e(n-1)+[X(n)+SM 0 ] (12)

[0108] Through the above derivation, the current weight deviation of the material is obtained, the mathematical model of the fully automatic water-washing packaging scale is established, and the fuzzy PID method is used for control.

[0109] The values ​​of the above formulas and contents are variables used to derive the mathematical model, but the final error expression does not include these, so they do not need to be measured in actual applications.

[0110] Step 2 is as follows:

[0111] The improved sand cat optimization algorithm optimizes the control parameters of PID. The basic sand cat optimization algorithm simulates the hunting behavior of sand cats, which mainly includes two aspects: searching and predation. The algorithm flow is as follows:

[0112] S1: First, initialize the population:

[0113] X i = l b +a×(u b -l b ) (13)

[0114] Where: X i represents the position of the i-th sand cat individual, a is a random number in [0,1], u b represents the upper bound, l b represents the lower bound; a is the variable for generating the initial population, so it is randomly selected between 0 and 1, and the upper and lower bounds are the limits of the search range. .

[0115] S2: Prey search phase:

[0116] The sand cat searches for prey by emitting low-frequency noise signals, and the signal perception range is 2kHz to 0. The intelligent optimization algorithm is a modeling algorithm that simulates animal habits. The data is animal habits, and the data is obtained from the public paper that proposed the algorithm. As the iteration proceeds, the signal value r G It decreases linearly from 2 to 0, simulating the process of a sand cat gradually approaching its prey without losing or jumping. M Imitates the auditory characteristics of the sand cat:

[0117]

[0118] t is the current iteration number, and T is the maximum iteration number. This value can be flexibly set according to different problems to be solved, which shows the universality of the proposed equation.

[0119] The main parameter controlling the transition between exploration and development process is R:

[0120] R = 2 × r G ×rand() (15)

[0121] Among them, rand() represents a random number between 0 and 1, r G Represents the process range where the sensitivity decreases linearly from 2 to 0.

[0122] In the search step, the update of each sand cat's current search position is based on a random position. This method allows the sand cat to explore new spaces in the search space. In order to prevent falling into the local optimum, each sand cat has a different sensitivity range r 2 :

[0123] r 2 =r G ×rand() (16)

[0124] Therefore, r 2 Used for position updates during exploration and development phases, r G It is used to guide the parameter R to switch between two stages.

[0125] In the sand cat optimization algorithm, the sand cat will update its position according to the optimal position of the current population, its own position and the sensitivity range. In order to ensure that the sand cat finds a new optimal position in the search area, which is between the current position and the target prey position, and at the same time ensure the convergence effect and complexity of the algorithm, the following position update mathematical model is established:

[0126]

[0127] where x t best Represents the position of the best individual in the current sand cat population.

[0128] S3: Predation stage:

[0129] The sand cat identifies prey based on the signals it hears. The position update model at this stage is as follows:

[0130]

[0131] Where rand() represents a random number between [0,1]. The search range of the sand cat is considered to be a circle. Its moving direction can be defined by a random angle θ on the circle, ranging from 0 to 360 degrees. Each sand cat in the population can move along a different circular direction to avoid falling into a local optimum. P is the distance the sand cat moves in one step, and x i t and x it+1 are the positions of the i-th sand cat at time t and time t+1 respectively.

[0132] Exploration and development of sand cat populations:

[0133] Exploration and development of sand cat colonies by R and r G The adaptive value of ensures that the Sand Cat Algorithm performs development (hunting) when the R value is less than or equal to 1, otherwise it can only perform exploration (search):

[0134]

[0135] Formula (19) represents the position update method of the i-th sand cat in the exploration and development stages. The different search methods of each sand cat prevent the algorithm from falling into the local optimum and ensure the global search capability.

[0136] Improvement point 1:

[0137] In step 2, Piecewise chaotic mapping is introduced to optimize the initialization of the sand cat population:

[0138]

[0139] Improvement 2:

[0140] In step 2, the Levy flight and walk strategy is introduced during the hunting process of the sand cat, which can provide a walking method with a step length that conforms to the Levy distribution. However, during the predation stage, the distance between the sand cat and the prey is very close. In order to be more consistent with the predation behavior of the sand cat, the adjustment constant C=0.37 is introduced to make the sand cat walk as close to the prey as possible. The Levy flight strategy formula is as follows:

[0141]

[0142] Levy is a strategy for updating data, and its representation is relatively complex;

[0143] Improvement point 3:

[0144] In the step 2, after the sand cat predation phase is over, a sparrow warning mechanism is introduced to prevent the sand cat from being discovered by its predator, thereby improving the convergence speed of the sand cat optimization algorithm.

[0145] When sand cats are aware of danger, those at the edge of the group will quickly move to a safe area to gain a better position, while those in the middle of the group will move randomly and get closer to other sand cats:

[0146]

[0147] Among them, x t best and x tworst Represent the current global best position and worst position respectively. β is a random number with controllable step size that obeys the normal distribution [0,1]. k represents the direction of the sand cat's movement and is a uniform random number in the range [-1,1]. and represent the current global best and worst fitness values ​​respectively, and ε is the minimum constant to avoid the denominator being 0. f i and f g Represent the current population fitness value and the global optimal fitness value, respectively, i =f g This means that the sand cat is on the edge of the population and is vulnerable to attacks by predators. It should be moved to a safe place as soon as possible.

[0148] Figure 2 The PID controller is optimized by the improved Sandcat optimization algorithm. The mathematical model of the fully automatic water-washing packaging scale is established to achieve precise control of the unloading module of the packaging scale, ensuring that the deviation between the input expected weight and the actual packaging weight is controlled within a reasonable range.

[0149] A fully automatic water-washing packaging scale unit based on ISCSO and its adaptive control system use basic test functions to verify the performance of the improved algorithm. The test function is selected from CEC2020, which is an authoritative test set of test functions. There are about 10 to 20 functions in it. It is used to test the performance of the algorithm, which is a single-objective optimization minimum function Hybrid Functions6. The improved sand cat algorithm is compared with the basic sand cat algorithm, particle swarm optimization algorithm, Harris eagle optimization algorithm, artificial gorilla troop optimization algorithm, sparrow search algorithm, and hippocampus optimization algorithm to verify the performance of the improved sand cat optimization algorithm. To ensure the fairness of the test, the population size of each algorithm is set to 50 and the maximum number of iterations is set to 500.

[0150] like Figure 4 As shown, it can be seen that the improved sand cat algorithm has greater advantages in convergence speed and optimization ability compared with other algorithms.

[0151] An automatic water-washing packaging scale unit based on ISCSO and its adaptive control system are used to verify the performance of the improved algorithm. The test function is selected from CEC2020 and is a single-objective optimization minimum function Composition Functions 4. The improved sand cat algorithm is compared with the basic sand cat algorithm, particle swarm optimization algorithm, Harris eagle optimization algorithm, artificial gorilla troop optimization algorithm, sparrow search algorithm, and hippocampus optimization algorithm to verify the performance of the improved sand cat optimization algorithm. To ensure the fairness of the test, the population size of each algorithm is set to 50 and the maximum number of iterations is set to 500.

[0152] like Figure 5As shown, it can be seen that the improved sand cat algorithm has greater advantages in convergence speed and optimization ability compared with other algorithms.

[0153] Figure 4 and Figure 5 "Convergence curve" is the "convergence curve", Dim is the number of dimensions, "Best scoreF*" is the name of the test function, the vertical axis represents the function value, and the horizontal axis is the number of iterations.

[0154] Working principle: The fully automatic water-washing packaging scale unit is taken as the research object. Its storage hopper receives the material put in by the unloading device, the load-bearing device receives and weighs the material quality, and returns the quality to the feeding device for material quality adjustment. The quantitative packaging device packages the material packages that meet the quality requirements. At the same time, the fully automatic water-washing device cleans the inside of the packaging scale through water and air spraying, preparing for the next weighing and packaging.

[0155] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An adaptive control method for a fully automatic water-washing packaging scale unit, the packaging scale unit to which the adaptive control method is applied comprises: The outer shell of the water-washing packaging scale (2), the storage hopper (1) installed on the upper part of the outer shell of the water-washing packaging scale (2), the full-automatic water-washing unit, the load-bearing unit (3) and the quantitative packaging unit (4); the full-automatic water-washing unit is integrated in the outer shell of the water-washing packaging scale (2), and the full-automatic water-washing unit includes a water / gas washing pipeline (6) and an arc-shaped diversion cleaning disc module; characterized in that the adaptive control method includes the following steps: Step 1: Analyze the feeding module of the full-automatic water-washing packaging scale, study the interaction relationship between feeding and weight, and establish a full-automatic water-washing packaging controller; Establish a mathematical model of the packaging scale according to the mechanical principle. Since it is impossible to accurately detect the magnitude I(t) of the impact force generated by the falling material at any moment and it is difficult to accurately predict the shutdown moment t, an online adjustment strategy is adopted; Assume that at the shutdown moment t, the measured weight of the nth bag of material is M(n), the falling material mass is Ml(n), the impact force generated by the falling material is I(n), the actual weight of the material is S(n), and the actual deviation at this time is e(n), and the nominal quantitative value is S. Then: S(n) = M(n) + Ml(n) - I(n) (1) e(n) = S - S(n) (2) e(n) = S - M(n) + [I(n) - Ml(n)] (3) Let X(n) = [I(n) - Ml(n)], X(n) represents the algebraic sum of the impact force of the falling material and the weight of the falling material in the air when weighing the nth bag of material, that is: e(n) = S - M(n) + X(n), n = 1, 2, 3,..., e(0) = 0 (4) Obviously, it can be known that: M(n) = M0 + e(n - 1) / 2 (5) Where M0 is a fixed value, and 2 / 3S < M0 < S, there is e(n) = S - M0 - e(n - 1) / 2 + X(n) (6) e(n - 1) = S - M0 - e(n - 2) / 2 + X(n - 1) (7) Subtract equation (7) from equation (6) to obtain the iterative formula: e(n) - e(n - 1) = 1 / 2[e(n - 1) - e(n - 2)] + [X(n) - X(n - 1)] (8) e(n) - e(1) = 1 / 2[e(n - 1) - e(0)] + [X(n) - X(1)] = -1 / 2e(n - 1) + [X(n) - X(1)] (9) e(n) = X(n) - 1 / 2e(n - 1) + e(1) - X(1) (10) From equation (1), e(1) = S - M(1) + X(1), then: e(n) = -1 / 2e(n - 1) + [X(n) + S - M(1)] (11) From equation (2), M(1) = M0, then: e(n) = -1 / 2e(n - 1) + [X(n) + S - M0] (12) Through the above derivation, the current weight deviation of the material is obtained, a mathematical model of the full-automatic water-washing packaging scale is established, and fuzzy PID control is adopted; Step 2: Design an improved sand cat optimization algorithm and introduce three improvement points; In the population initialization stage, the Piecewise chaos mapping theory is introduced to initialize the sand cat population; in the attack stage, the Levy flying and wandering strategy is introduced; since the distance between the sand cat and the prey is very close when attacking the prey, in order to better conform to the behavior of the sand cat attacking the prey, the adjustment coefficient C=0.37 is introduced to make the sand cat wander near the prey; Specifically: The improved sand cat optimization algorithm optimizes the control parameters of PID. The basic sand cat optimization algorithm simulates the hunting behavior of sand cats, which mainly includes two aspects: searching and predation. The algorithm flow is as follows: First, initialize the population: X i =l b +a×(u b -l b ) (13) Where: X i represents the position of the i-th sand cat individual, a is a random number in [0,1], u b represents the upper bound, l b Indicates the lower bound; Prey Search Phase: The sand cat searches for prey by emitting low-frequency noise signals. The signal sensing range is from 2kHz to 0. As the iteration proceeds, the signal value r G It decreases linearly from 2 to 0, simulating the process of a sand cat gradually approaching its prey without losing or jumping. M Imitates the auditory characteristics of the sand cat: t is the current iteration number, T is the maximum iteration number, and this value can be flexibly set according to different problems to be solved, which shows the universality of the proposed equation; The main parameter controlling the transition between exploration and development process is R: R=2×r G ×rand() (15) Among them, rand() represents a random number between 0 and 1, r G Indicates the process range in which the sensitivity decreases linearly from 2 to 0; In the search step, the update of each sand cat's current search position is based on a random position. This method allows the sand cat to explore new space in the search space. In order to prevent falling into the local optimum, each sand cat has a different sensitivity range r2: r2=r G ×rand() (16) Therefore, r2 is used to update the position during the exploration and development phases, and r G It is used to guide the parameter R to switch between two stages; In the sand cat optimization algorithm, the sand cat will update its position according to the optimal position of the current population, its own position and the sensitivity range. In order to ensure that the sand cat finds a new optimal position in the search area, which is between the current position and the target prey position, and at the same time ensure the convergence effect and complexity of the algorithm, the following position update mathematical model is established: where x t best Represents the position of the best individual in the current sand cat group; Predation stage: The sand cat identifies prey based on the signals it hears. The position update model at this stage is as follows: Where rand() represents a random number between [0,1]. The search range of the sand cat is considered to be a circle. Its moving direction is defined by a random angle θ on the circle, ranging from 0 to 360 degrees. Each sand cat in the population moves along a different circular direction to avoid falling into a local optimum. P is the distance the sand cat moves in one step, and x i t and x i t+1 are the positions of the i-th sand cat at time t and time t+1 respectively; Exploration and development of sand cat populations: Exploration and development of sand cat colonies by R and r G The adaptive value of ensures that the Sand Cat Algorithm performs development (hunting) when the R value is less than or equal to 1, otherwise it can only perform exploration (search): Formula (19) represents the position update method of the i-th sand cat in the exploration and development stages. The different search methods of each sand cat prevent the algorithm from falling into the local optimum and ensure the global search capability. Piecewise chaotic mapping is introduced to optimize the initialization of sand cat population: The Levy flight and wandering strategy is introduced in the hunting process of the sand cat to provide a wandering method with a step length that conforms to the Levy distribution; in the predation stage, the distance between the sand cat and the prey is very close. In order to be more consistent with the predation behavior of the sand cat, the adjustment constant C=0.37 is introduced to make the sand cat wander near the prey. The Levy flight strategy formula is as follows: Step 3: After the sand cat hunting phase is over, a sparrow alert mechanism is introduced. When the sand cat group realizes the danger, the edge sand cats quickly move to a safe area, and the sand cats in the population randomly move close to other sand cats to prevent the sand cats from being discovered by their predators, thereby improving the convergence speed of the sand cat optimization algorithm. When sand cats are aware of danger, those at the edge of the group will quickly move to a safe area to gain a better position, while those in the middle of the group will move randomly and get closer to other sand cats: Among them, x t best and x t worst They represent the current global best position and the worst position respectively; β is a random number with a controllable step size that obeys the normal distribution of [0,1], k represents the direction of the sand cat's movement, which is a uniform random number in the range of [-1,1]; and represent the current global best and worst fitness values ​​respectively, and ε is the minimum constant to avoid the denominator being 0; when f i =f g This means that the sand cat is on the edge of the population and is vulnerable to attacks by predators. At this time, it should be moved to a safe place as soon as possible.

Images