Latex mattress production device and process
By using a motor-driven mixing tank and heating mixing rack in the latex mattress production device, combining intelligent sensors and genetic algorithms to optimize process parameters, the temperature problem in latex mattress production is solved, and the uniform mixing and rapid cooling of latex is achieved. Improve product quality and production efficiency.
Patent Information
- Application Number
- CN202410854848.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-27
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The latex mattress is viscous when the temperature is low during the production process, resulting in uneven stirring, affecting elasticity and comfort, and the cooling process is too slow, extending the production cycle and affecting the long-term use stability and durability of the product.
A latex mattress production device was designed, including a motor-driven mixing tank and a heating mixing rack. The thermally conductive copper tube is used to maintain the latex temperature uniformity, combine intelligent sensors and predictive control algorithms to adjust the mold temperature and pressure in real time, and optimize the extrusion through genetic algorithms. Press the process parameters to ensure the consistency of product hardness and thickness.
It improves the uniformity and fluidity of latex mixing, shortens production time, improves product quality and production efficiency, and ensures the stability and durability of latex mattresses.
Smart Images

Figure CN118990914A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of latex mattresses, and in particular relates to a latex mattress production device and a process thereof. Background Art
[0002] Latex mattress is a modern green bedroom product. The latex sponge in the latex mattress is a porous rubber material. Its pore structure is mainly all connected holes or most of the connected holes, and a small part is not connected holes. This structure gives the latex mattress the characteristics of high elasticity, shock absorption, compression fatigue resistance, good load-bearing capacity, comfort and durability. The high elasticity of the latex mattress can meet the needs of people of different weights, and its good support can adapt to various sleeping positions of sleepers. In the production equipment of latex mattresses, the processing of raw materials is a particularly important link.
[0003] In the prior art, the authorization publication number "CN110640936A" discloses a "latex slurry impurity removal device for latex mattress production"; it includes an outer shell cylinder, a third through hole is opened on the left side of the outer shell cylinder, a rotating rod is rotatably installed on the bottom surface of the inner cavity of the outer shell cylinder, a second driven gear is fixedly sleeved on the lower part of the surface of the rotating rod, a second motor is fixedly installed on the right side of the bottom surface of the inner cavity of the outer shell cylinder, and a second driving gear meshing with the second driven gear is fixedly installed on the output end of the second motor, a baffle is fixedly inserted on the lower part of the inner ring surface of the outer shell cylinder, and the baffle is sleeved on the lower part of the rotating rod, a thread cutter is fixedly sleeved on the middle part of the surface of the rotating rod, and a blade is fixedly sleeved on the upper part of the surface of the rotating rod. By adding a rotating rod, a thread cutter, a blade, a fixing ring, a fixing cylinder and a wide mouth mask to the outer shell cylinder, the blade cuts large pieces of oak bark into small pieces, and the thread cutter cuts small pieces of oak bark into pieces, thereby preventing large pieces of oak bark from being stuck in the spiral leaves.
[0004] The above-mentioned "latex slurry impurity removal device for latex mattress production" still has some disadvantages, for example: latex needs to be stirred during the production process, and if the temperature is low during the stirring process, the latex will become viscous, resulting in uneven latex during the stirring process, and it cannot be fully foamed, affecting the elasticity and comfort of the subsequent processing of the latex mattress, and the density and hardness of the mattress are uneven. At the same time, the latex mattress needs to be cooled and solidified after stirring to form a stable structure. The existing cooling process is too slow, which will extend the entire production cycle and reduce production efficiency. In addition, the too slow cooling process will cause stress inside the latex mattress, affecting its stability and durability in long-term use;
[0005] For this reason, a latex mattress production device and process thereof are proposed here to solve the above-mentioned problems. Summary of the invention
[0006] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a latex mattress production device and a process thereof, which effectively solves the problem that when the temperature is low, the latex becomes viscous, resulting in uneven latex during the stirring process and failure to fully foam, affecting the elasticity and comfort of the subsequent processed latex mattress, and causing uneven density and hardness of the mattress. At the same time, the latex mattress needs to be cooled and solidified after stirring to form a stable structure. The existing cooling process is too slow, which will extend the entire production cycle and reduce production efficiency. In addition, the too slow cooling process will cause stress inside the latex mattress, affecting its stability and durability during long-term use.
[0007] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a latex mattress production device, comprising a processing base, a mixing tank is arranged on the top of the processing base, a top cover is fixedly connected to the top of the mixing tank, a motor body is fixedly connected to the surface of the processing base, an output shaft of the motor body is movably connected to an active top tooth, one side of the active top tooth is movably connected to a transmission side tooth, and the top of the transmission side tooth is fixedly connected to the bottom of the mixing tank.
[0008] Preferably: the top of the active top tooth is fixedly connected to the bottom transmission rod, the top of the bottom transmission rod is fixedly connected to the top active toothed disc, one side of the top active toothed disc is movably connected to the side transmission toothed disc, one side of the side transmission toothed disc is fixedly connected to the side transmission rod, one end of the side transmission rod is fixedly connected to the side movable toothed disc, one side of the side movable toothed disc is movably connected to the side connecting toothed disc, the inner side of the side connecting toothed disc is fixedly connected to the bottom connecting rod, the surface of the bottom connecting rod is fixedly connected to the top limiting hole, and one side of the top limiting hole is fixedly connected to the stirring frame.
[0009] Preferably, a heating box is provided on the inner side of the stirring frame, a heating tube is movably connected to the inner side of the heating box, a transmission cable is provided on the inner side of the top limit hole, and a connecting groove is provided on the inner side of the transmission cable.
[0010] Preferably: the top of the processing base is fixedly connected to a side support frame, one side of the side support frame is fixedly connected to a fixing frame, a top connecting hole is opened on the top of the fixing frame, and the inner side of the top connecting hole is arranged on the surface of the bottom connecting rod.
[0011] Preferably, a top connecting frame is fixedly connected to the top of the fixing frame, a battery body is fixedly connected to the top of the top connecting frame, and a controller body is arranged on one side of the battery body.
[0012] Preferably: a cooling circulation pipe is opened on the inner side of the mixing tank, and measuring liquid storage boxes are fixedly connected to both sides of the mixing tank. A liquid inlet is provided on the top of the measuring liquid storage box, and a water pump body is provided on one side of the liquid inlet. A top drainage pipe is fixedly connected to the other side of the measuring liquid storage box, and a bottom drainage pipe is provided at the bottom of the measuring liquid storage box.
[0013] Preferably, a heat-conducting copper tube is fixedly connected to one side of the stirring frame, and there are multiple heat-conducting copper tubes.
[0014] Preferably, a top limiting hole is opened on the surface of the top cover, and the inner side of the top limiting hole is arranged on the surface of the bottom connecting rod.
[0015] Preferably, the bottom of the transmission side teeth is movably connected to a bearing disk, the bottom of the bearing disk is movably connected to a bottom support platform, and the bottom of the bottom support platform is fixedly connected to the top of the processing base.
[0016] A latex mattress production process comprises the following steps:
[0017] S1. Prepare latex and additives, and prepare the latex and additives according to the formula ratio;
[0018] S2, adding the raw materials of latex and additives into a mixing tank for mixing and stirring;
[0019] S3, putting the mixed raw materials into a foaming machine, stirring at high speed, adding an appropriate amount of foaming agent, and then making the latex react chemically in the mold by high temperature and pressure to form latex foam;
[0020] S4. Inject the latex foam into the pre-prepared mold to automatically shape it. The mold is made according to the size and shape of the mattress to ensure the precise size of the mattress;
[0021] S5. The latex mattress after forming needs to go through the extrusion process to achieve the appropriate hardness and thickness, and the latex mattress after extrusion needs to be trimmed to remove the burrs and uneven parts on the surface;
[0022] S6. Clean the processed latex mattress to remove excess residue and odor;
[0023] S7. Conduct quality inspection on the produced latex mattresses to ensure that the products meet the relevant standards and requirements;
[0024] S8. Package the repaired latex mattress to prevent damage during transportation.
[0025] According to S4 and S5, smart sensors are installed in the mold, and the predictive control algorithm LSTM is used to predict the molding process of latex foam in the mold in real time, and the temperature and pressure of the mold are dynamically adjusted to ensure the precise size and shape of the mattress. Then, by installing force sensors and displacement sensors, the extrusion process parameters pressure and extrusion speed are optimized using genetic algorithms to ensure that the latex mattress reaches the appropriate hardness and thickness, while reducing the defective rate. The specific process is:
[0026] Step 1: Data collection and preprocessing
[0027] Smart sensor installation
[0028] Pressure sensors: Installed at multiple key locations of the mold to monitor pressure distribution throughout the molding process.
[0029] Temperature sensor: installed at different locations inside and outside the mold to monitor the temperature changes of latex foam.
[0030] Force sensor: installed at the key stress-bearing parts of the extrusion equipment to monitor the force changes during the entire extrusion process.
[0031] Displacement sensor: installed on the moving parts of the extrusion equipment to monitor the displacement changes during the extrusion process in real time.
[0032] Real-time data collection:
[0033] Sampling frequency: 100 times per second.
[0034] Data storage: Use PLC programmable logic controller for preliminary data processing and storage, and then transmit it to the central control system for analysis.
[0035] Data preprocessing:
[0036] Data cleaning: Remove outliers and noise that may appear during the collection process, and use a moving average filter for smoothing.
[0037] Data standardization: Standardize the pressure and temperature data to facilitate subsequent machine learning algorithm processing. Standardization formula:
[0038]
[0039] Among them, v is the original data, μ is the mean of the data, and e is the standard deviation of the data.
[0040] Step 2: Model building and training
[0041] A prediction model for the hardness and thickness of latex mattresses was established and trained using the machine learning algorithm LSTM. In order to improve the accuracy of the prediction model for the hardness and thickness of latex mattresses, the following improvements can be made to the standard LSTM:
[0042] Bidirectional LSTM: Bidirectional LSTM takes into account both forward and backward information to enhance the modeling capability of time series data; it consists of two LSTM layers, one for processing the forward sequence and the other for processing the reverse sequence. The specific formula is as follows:
[0043] Forward LSTM formula:
[0044] f t =σ(W f ·[h t-1 , x t ]+b f ),
[0045] i t =σ(W i ·[h t-1 , x t ]+b i ),
[0046]
[0047] o t =σ(W o ·[h t-1 , x t ]+b o )
[0048] h t =o t ·tan h(C t )
[0049] Among them, f t is the activation vector of the forget gate, which determines the cell state C t-1 The degree to which it needs to be forgotten; t is the activation vector of the input gate, which determines the new information to what extent the cell state is updated; is the candidate cell state, containing new candidate information; C t is the current cell state, which is obtained by weighted sum of the cell state at the previous moment and the current candidate state. t is the activation vector of the output gate, which determines the cell state C t To what extent does it affect the hidden state h t ;h t is the current hidden state, through the cell state C t Weighted; W f , W i , W C , W o is the weight matrix, which is used to calculate the forget gate, input gate, candidate cell state and output gate respectively; b f 、bi 、b C 、b o is the bias term, which is used for the calculation of the forget gate, input gate, candidate cell state and output gate respectively; σ is the activation function; tanh is the hyperbolic tangent activation function; x t is the current input vector; h t-1 is the hidden state at the previous moment.
[0050] Reverse LSTM formula:
[0051] f′ t =σ(W′ f ·[h′ t+1 , x t ]+b′ f )
[0052] i′ t =σ(W′ i ·[h′ t+1 , x t ]+b′ i )
[0053]
[0054] o′ t =σ(W′ o ·[h′ t+1 , x t ]+b′ o )
[0055] h′ t =o′ t tanh(C′ t )
[0056] The meaning of the symbols in the reverse LSTM formula is the same as that in the forward LSTM formula. “′” represents the parameter in the reverse sequence.
[0057] Bidirectional LSTM output:
[0058] H t =[h t ; h′ t ],
[0059] Among them, H t is the bidirectional LSTM output at the current moment, which is composed of the hidden state h of the forward LSTM t and the hidden state h′ of the reverse LSTM t Splicing
[0060] Attention mechanism: Add attention mechanism to highlight the impact of data at key time steps on prediction.
[0061] Attention weights:
[0062] e t =v T tanh(W H H t +b H ),
[0063] Among them, e t is the attention score at the current moment, indicating the importance of time step ttt; v T Weight vector, used to calculate the attention score; W H is the weight matrix of the attention layer; b H is the bias of the attention layer; H t is the bidirectional LSTM output at the current moment.
[0064] Attention score:
[0065]
[0066] Among them, α t is the attention weight at the current moment, indicating the normalized importance of time step t; e t is the attention score at the current moment; T is the total length of the sequence.
[0067] Context vector:
[0068]
[0069] Where c is the context vector, which is obtained by weighted summing the attention weights and bidirectional LSTM outputs at all time steps; α t is the attention weight at the current moment. t is the bidirectional LSTM output at the current moment.
[0070] Multi-task learning: Predict hardness and thickness simultaneously and improve prediction accuracy through shared feature representation.
[0071] Hardness prediction:
[0072]
[0073] in, is the predicted hardness value; W hardness is the weight matrix of the hardness prediction layer. b hardness is the bias of the hardness prediction layer. c is the context vector.
[0074] Thickness prediction:
[0075]
[0076] in, is the predicted thickness value. Wthickness is the weight matrix of the thickness prediction layer. W thickness is the bias of the thickness prediction layer and c is the context vector.
[0077] Loss function formula:
[0078] Hardness loss:
[0079]
[0080] Among them, L hardness Loss function for hardness prediction; N is the number of samples; The predicted hardness value of the i-th sample; y hardness,i is the true hardness value of the i-th sample.
[0081] Thickness loss:
[0082]
[0083] Among them, L thickness Loss function for thickness prediction; N is the number of samples; is the predicted thickness value of the ith sample. thickness,i is the true thickness value of the i-th sample.
[0084] Total loss:
[0085] L total =λ 1 L hardness +λ 2 L thickness ,
[0086] Among them, L total is the total loss function used to optimize hardness and thickness prediction simultaneously; λ 1 and λ 2 are the weight coefficients of hardness and thickness loss, respectively, controlling their contributions to the total loss.
[0087] Step 3: Optimization algorithm application
[0088] Genetic algorithm is used to iteratively find the best combination of extrusion process parameters. In order to improve the efficiency and accuracy of extrusion process parameter optimization, the following improvements can be made to the standard genetic algorithm:
[0089] Adaptive crossover and mutation: Adaptively adjust the crossover rate and mutation rate according to fitness.
[0090] An initial population is randomly generated, each individual represents a combination of extrusion process parameters:
[0091] Population size: Set the population size to P.
[0092] Individual coding: Each individual x is encoded with a set of parameters x = (p 1 , p 2 , p 3 …p n ) indicates that p i is the extrusion process parameter.
[0093] Fitness function:
[0094] Define the fitness function f(x) to evaluate the quality of each individual, usually calculated based on hardness and thickness errors:
[0095] f(x)=w 1 ·hardness error (x)+w 2 ·thickness error (x)
[0096] Among them, hardness error (x) and thickness error (x) represents the hardness and thickness error of individual x respectively. 1 and w 2 is the weight coefficient of hardness and thickness error.
[0097] Multiple selection strategy combinations: Combine roulette selection and tournament selection, taking into account both exploration and development.
[0098] Roulette wheel selection: individuals are selected with probability proportional to their fitness values.
[0099] Tournament selection: randomly select k individuals from the population and choose the one with the best fitness value as the parent.
[0100] Adaptive crossover rate is used to adjust the crossover rate according to the fitness difference of parents:
[0101]
[0102] Among them, α is the crossover rate; α max and α min are the minimum and maximum crossover rates respectively; Δf is the fitness difference between the parents. max and f min are the maximum and minimum fitness values in the current population respectively.
[0103] Single point crossover:
[0104] x new1 =α·x parent1 +(1-α)·x parent2
[0105] x new2 =(1-α)·xparent1 +α·x parent2
[0106] Among them, x new1 is a new individual 1 generated by crossover; x new2 is the new individual 2 generated by crossover; α is the crossover rate; x parent1 is the parent individual 1; x parent2 is parent individual 2;
[0107] Mutation: Adaptive mutation rate is used to adjust the mutation rate according to individual fitness.
[0108]
[0109] Among them, β is the mutation rate; β max and β min are the minimum and maximum mutation rates respectively; f(x) is the fitness value of the current individual; f max and f min are the maximum and minimum fitness values in the current population respectively.
[0110] Mutation operation:
[0111] x new =x+β·δ,
[0112] Among them, β is the mutation rate; δ is the random disturbance, which obeys the standard normal distribution; x new is the new individual after mutation; x is the original individual;
[0113] Elite retention strategy: retain the best E individuals in the previous generation and directly enter the next generation to avoid the loss of excellent genes.
[0114] Elite individuals: Select E individuals with the highest fitness values from the previous generation and directly add them to the new population.
[0115] Local search: Perform local search on the current optimal solution during the iteration process to improve the accuracy of the solution:
[0116] x′ new =x best +γ·δ,
[0117] Among them, x′ new is the new individual after local search; x best is the current optimal individual; γ is the local search step length; δ is a random disturbance, which usually obeys the standard normal distribution;
[0118] Iterative optimization: Repeat selection, crossover, mutation, elite retention, and local search operations until a stopping condition is reached.
[0119] Step 4: Real-time optimization algorithm
[0120] According to the optimization results, the parameters of the extrusion equipment are adjusted in real time to ensure that the hardness and thickness of the mattress meet the requirements.
[0121] Real-time optimization and control:
[0122] Real-time fitness calculation and parameter adjustment: Calculate the fitness value of the current extrusion process parameters in real time, and perform crossover and mutation operations to generate new parameter combinations.
[0123] According to the optimization results, the parameters of the extrusion equipment are adjusted in real time: the target extrusion force is 15kN, which is achieved by adjusting the hydraulic system. The target extrusion speed is 50mm / s, and the motor speed is controlled by the inverter.
[0124] Feedback control system:
[0125] Use PID controller to provide real-time feedback and correction of adjusted parameters:
[0126]
[0127] Among them, u(t) is the control input hydraulic valve opening and motor voltage, e(t) is the error, K p , K i , K d are the proportional, integral, and differential coefficients.
[0128] Real-time monitoring and alarm system:
[0129] Real-time monitoring system: A monitoring system is installed in the central control room to display graphs and data of squeeze force, displacement, mattress hardness and thickness in real time. Set alarm thresholds, and when parameters exceed the set range, an alarm signal is automatically issued to prompt the operator to intervene.
[0130] Alarm processing: When the system detects an abnormal situation, such as the extrusion pressure is too high or too low, or the extrusion speed is too fast or too slow, it will automatically trigger an alarm and record the abnormal data. The operator will adjust the equipment in time according to the alarm information to ensure the stability of the production process.
[0131] Data storage and analysis:
[0132] Data storage: The real-time collected data is stored in the database for subsequent analysis and optimization. The stored data includes the extrusion force, displacement, hardness, thickness and corresponding optimization parameters at each time step.
[0133] Data analysis: Regularly analyze the stored data to evaluate the stability of the production process and product quality. Use data mining and machine learning techniques to identify problems in the production process and propose improvement plans.
[0134] Compared with the prior art, the present invention has the following beneficial effects:
[0135] 1) In the latex mattress production device and its process, after the motor body is started, the motor body drives the active top gear to rotate, and through the meshing of the active top gear and the transmission side gear, the rotating active top gear drives the transmission side gear to rotate, and the rotating mixing tank is used to mix the latex inside, which helps to improve the uniformity of latex mixing and the fluidity of latex, thereby improving the overall quality and performance of the mattress;
[0136] 2) In the latex mattress production device and its process, the rotation direction of the bottom connecting rod and the active top gear can be changed through the transmission connection of the top active gear plate, the side transmission gear plate, the side transmission rod, the side movable gear plate and the side connecting gear plate. At this time, when the mixing tank rotates to mix the latex, the stirring frame will rotate in the opposite direction of the mixing tank to stir the latex, which can improve the stirring effect of the latex, improve the mixing property of the latex, reduce the mixing time, and improve the mixing efficiency;
[0137] 3) In the latex mattress production device and its process, the controller body drives the heating tube to work, and the controller body penetrates the connecting groove opened on the inner side of the top limit hole and enters the heating box opened on the inner side of the stirring frame. By driving the heating tube to work, the temperature in the heating box can be increased, and the increased temperature will be transmitted to the latex through the heat-conducting copper tube, so that the temperature of the latex can be continuously increased by the heat-conducting copper tube during the stirring of the latex when the stirring frame rotates, which can avoid the phenomenon of the latex gradually becoming sticky and improve the fluidity and effect of the latex;
[0138] 4) In the latex mattress production device and its process, the temperature of the mixing tank can be quickly lowered by circulating the coolant in the mixing tank, so that the latex processed in the mixing tank can be cooled, and the circulating coolant can flow evenly along the inner wall of the mixing tank, which can improve the uniformity of cooling and improve the efficiency and effect of latex cooling.
[0139] 5) The improved LSTM model and the improved genetic algorithm are combined and applied to the latex mattress production device, which significantly improves the intelligent level of the production process, product quality and production efficiency. The improved LSTM model is used to adjust the temperature and pressure of the mold in real time to ensure the stability of the thickness and hardness of the latex mattress, and dynamically adjust the production parameters based on the real-time prediction of the thickness and hardness of the latex mattress; the improved genetic algorithm is used to optimize the extrusion force and extrusion speed to ensure that the product reaches the optimal hardness and thickness, reduce the defective rate, save energy and optimize resource utilization, reduce production costs, and improve product consistency and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0140] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0141] Figure 1 It is a schematic diagram of the overall appearance structure of the present invention;
[0142] Figure 2 It is a schematic diagram of the top cover structure of the present invention;
[0143] Figure 3 It is a schematic diagram of the bottom transmission rod structure of the present invention;
[0144] Figure 4 It is a schematic diagram of the bottom connecting rod structure of the present invention;
[0145] Figure 5 It is a schematic diagram of the connecting groove structure of the present invention;
[0146] Figure 6 It is a schematic diagram of the water pump body structure of the present invention.
[0147] Figure 7 Intelligent optimization flow chart for the molding process and extrusion process
[0148] In the figure: 1, processing base; 2, mixing tank; 3, top cover; 401, motor body; 402, driving top gear; 403, transmission side gear; 404, bearing plate; 405, bottom support platform; 501, bottom transmission rod; 502, top driving gear plate; 503, side transmission gear plate; 504, side transmission rod; 505, side movable gear plate; 506, side connecting gear plate; 507, bottom connecting rod; 508, top limit hole; 509, stirring frame; 5 010, fixing frame; 5011, side support frame; 5012, top connecting hole; 601, top connecting frame; 602, controller body; 603, battery body; 604, transmission cable; 605, connecting groove; 606, heating box; 607, heating pipe; 608, heat-conducting copper pipe; 701, liquid storage box; 702, liquid inlet; 703, water pump body; 704, cooling circulation pipe; 705, top drainage pipe; 706, bottom drainage pipe. DETAILED DESCRIPTION
[0149] 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.
[0150] Example 1
[0151] In this embodiment, Figure 1-Figure 6 Given, the present invention provides the following technical solutions:
[0152] A latex mattress production device includes a processing base 1, a mixing tank 2 is arranged on the top of the processing base 1, a top cover 3 is fixedly connected to the top of the mixing tank 2, a motor body 401 is fixedly connected to the surface of the processing base 1, an output shaft of the motor body 401 is movably connected to an active top tooth 402, one side of the active top tooth 402 is movably connected to a transmission side tooth 403, and the top of the transmission side tooth 403 is fixedly connected to the bottom of the mixing tank 2.
[0153] It should be noted that, through the meshing connection between the transmission side teeth 403 and the active top teeth 402 , when the active top teeth 402 are driven to rotate by the motor body 401 , the rotating active top teeth 402 will drive the transmission side teeth 403 to rotate.
[0154] In this embodiment, the top of the active top tooth 402 is fixedly connected to the bottom transmission rod 501, the top of the bottom transmission rod 501 is fixedly connected to the top active toothed disc 502, one side of the top active toothed disc 502 is movably connected to the side transmission toothed disc 503, one side of the side transmission toothed disc 503 is fixedly connected to the side transmission rod 504, one end of the side transmission rod 504 is fixedly connected to the side movable toothed disc 505, one side of the side movable toothed disc 505 is movably connected to the side connecting toothed disc 506, the inner side of the side connecting toothed disc 506 is fixedly connected to the bottom connecting rod 507, the surface of the bottom connecting rod 507 is fixedly connected to the top limiting hole 508, and one side of the top limiting hole 508 is fixedly connected to the stirring frame 509.
[0155] It should be noted that the two top limit holes 508 will respectively drive the stirring frame 509 to rotate and stir along the inner side of the mixing tank 2, and at this time, through the transmission connection of the top active gear plate 502, the side transmission gear plate 503, the side transmission rod 504, the side movable gear plate 505 and the side connecting gear plate 506, the rotation direction of the bottom connecting rod 507 and the active top tooth 402 can be changed. At this time, when the mixing tank 2 rotates to mix the latex, the stirring frame 509 will rotate in the opposite direction of the mixing tank 2 to stir the latex.
[0156] In this embodiment, a heating box 606 is opened on the inner side of the stirring frame 509, a heating tube 607 is movably connected to the inner side of the heating box 606, a transmission cable 604 is opened on the inner side of the top limit hole 508, and a connecting groove 605 is provided on the inner side of the transmission cable 604.
[0157] It should be noted that the controller body 602 passes through the connecting groove 605 opened on the inner side of the top limit hole 508 and enters the heating box 606 opened on the inner side of the stirring frame 509. By driving the heating tube 607 to work, the temperature in the heating box 606 can be increased.
[0158] In this embodiment, a side support frame 5011 is fixedly connected to the top of the processing base 1, a fixing frame 5010 is fixedly connected to one side of the side support frame 5011, a top connecting hole 5012 is opened at the top of the fixing frame 5010, and the inner side of the top connecting hole 5012 is arranged on the surface of the bottom connecting rod 507.
[0159] It should be noted that, by providing the top communicating hole 5012 , the rotation stability of the bottom connecting rod 507 can be maintained, thereby preventing the bottom connecting rod 507 from being offset during the rotation process.
[0160] In this embodiment, a top connecting frame 601 is fixedly connected to the top of the fixing frame 5010 , a battery body 603 is fixedly connected to the top of the top connecting frame 601 , and a controller body 602 is disposed on one side of the battery body 603 .
[0161] It should be noted that after the battery body 603 is connected to an external power source, the controller body 602 can be driven to work by the battery body 603 .
[0162] In this embodiment, a cooling circulation pipe 704 is opened on the inner side of the mixing tank 2, and measuring liquid storage tanks 701 are fixedly connected on both sides of the mixing tank 2. A liquid inlet 702 is provided on the top of the measuring liquid storage tank 701, and a water pump body 703 is provided on one side of the liquid inlet 702. A top drainage pipe 705 is fixedly connected to the other side of the measuring liquid storage tank 701, and a bottom drainage pipe 706 is provided at the bottom of the measuring liquid storage tank 701.
[0163] It should be noted that the coolant enters the liquid storage tank 701 through the bottom drain pipe 706 to form a circulation. The temperature of the mixing tank 2 can be quickly lowered by the coolant circulating in the mixing tank 2, so that the latex processed in the mixing tank 2 can be cooled. The circulating coolant can flow evenly along the inner wall of the mixing tank 2, which can improve the uniformity of cooling and improve the efficiency and effect of latex cooling.
[0164] In this embodiment, a heat-conducting copper tube 608 is fixedly connected to one side of the stirring frame 509 , and there are multiple heat-conducting copper tubes 608 .
[0165] It should be noted that by driving the heating tube 607 to work, the temperature in the heating box 606 can be increased, and the increased temperature will be transmitted to the latex through the heat-conducting copper tube 608, so that the temperature of the latex can be continuously increased by using the heat-conducting copper tube 608 while the latex is stirred when the stirring frame 509 rotates.
[0166] In this embodiment, a top limiting hole 508 is opened on the surface of the top cover 3 , and the inner side of the top limiting hole 508 is arranged on the surface of the bottom connecting rod 507 .
[0167] It should be noted that when the bottom connecting rod 507 rotates, it will drive the two top limiting holes 508 to rotate, and the two top limiting holes 508 will respectively drive the stirring frame 509 to rotate and stir along the inner side of the mixing tank 2.
[0168] In this embodiment, the bottom of the transmission side gear 403 is movably connected to a bearing plate 404 , the bottom of the bearing plate 404 is movably connected to a bottom support platform 405 , and the bottom of the bottom support platform 405 is fixedly connected to the top of the processing base 1 .
[0169] It should be noted that the bearing plate 404 is movably connected to the surface of the bottom support platform 405 for support, and the connection of the bearing plate 404 can improve the stability of the transmission side teeth 403 during rotation.
[0170] A latex mattress production device and process thereof, comprising the following steps:
[0171] S1. Prepare latex and additives, and prepare the latex and additives according to the formula ratio;
[0172] S2, adding the raw materials of latex and additives into mixing tank 2 for mixing and stirring;
[0173] S3, putting the mixed raw materials into a foaming machine, stirring at high speed, adding an appropriate amount of foaming agent, and then making the latex react chemically in the mold by high temperature and pressure to form latex foam;
[0174] S4. Inject the latex foam into the pre-prepared mold to automatically shape it. The mold is made according to the size and shape of the mattress to ensure the precise size of the mattress;
[0175] S5. The latex mattress after forming needs to go through the extrusion process to achieve the appropriate hardness and thickness, and the latex mattress after extrusion needs to be trimmed to remove the burrs and uneven parts on the surface;
[0176] S6. Clean the processed latex mattress to remove excess residue and odor;
[0177] S7. Conduct quality inspection on the produced latex mattresses to ensure that the products meet the relevant standards and requirements;
[0178] S8. Package the repaired latex mattress to prevent damage during transportation.
[0179] According to S4 and S5, smart sensors are installed in the mold, and the predictive control algorithm LSTM is used to predict the molding process of latex foam in the mold in real time, and the temperature and pressure of the mold are dynamically adjusted to ensure the precise size and shape of the mattress. Then, by installing force sensors and displacement sensors, the extrusion process parameters pressure and extrusion speed are optimized using genetic algorithms to ensure that the latex mattress reaches the appropriate hardness and thickness, while reducing the defective rate. The specific process is:
[0180] Step 1: Data collection and preprocessing
[0181] Smart sensor installation
[0182] Pressure sensors: Installed at multiple key locations of the mold to monitor pressure distribution throughout the molding process.
[0183] Temperature sensor: installed at different locations inside and outside the mold to monitor the temperature changes of latex foam.
[0184] Force sensor: installed at the key stress-bearing parts of the extrusion equipment to monitor the force changes during the entire extrusion process.
[0185] Displacement sensor: installed on the moving parts of the extrusion equipment to monitor the displacement changes during the extrusion process in real time.
[0186] Real-time data collection:
[0187] Sampling frequency: 100 times per second.
[0188] Data storage: Use PLC programmable logic controller for preliminary data processing and storage, and then transmit it to the central control system for analysis.
[0189] Data preprocessing:
[0190] Data cleaning: Remove outliers and noise that may appear during the collection process, and use a moving average filter for smoothing.
[0191] Data standardization: Standardize the pressure and temperature data to facilitate subsequent machine learning algorithm processing. Standardization formula:
[0192]
[0193] Among them, v is the original data, μ is the mean of the data, and e is the standard deviation of the data.
[0194] Step 2: Model building and training
[0195] A prediction model for the hardness and thickness of latex mattresses was established and trained using the machine learning algorithm LSTM. In order to improve the accuracy of the prediction model for the hardness and thickness of latex mattresses, the following improvements can be made to the standard LSTM:
[0196] Bidirectional LSTM: Bidirectional LSTM takes into account both forward and backward information to enhance the modeling capability of time series data; it consists of two LSTM layers, one for processing the forward sequence and the other for processing the reverse sequence. The specific formula is as follows:
[0197] Forward LSTM formula:
[0198] f t =σ(W f ·[h t-1 , x t ]+b f ),
[0199] i t =σ(W i ·[h t-1 , x t ]+b i ),
[0200]
[0201] o t =σ(W o ·[h t-1 , x t ]+b o )
[0202] h t =o t ·tan h(C t )
[0203] Among them, f t is the activation vector of the forget gate, which determines the cell state C t-1 The degree to which it needs to be forgotten; t is the activation vector of the input gate, which determines the new information to what extent the cell state is updated; is the candidate cell state, containing new candidate information; C t is the current cell state, which is obtained by weighted sum of the cell state at the previous moment and the current candidate state. t is the activation vector of the output gate, which determines the cell state C t To what extent does it affect the hidden state h t ;h t is the current hidden state, through the cell state C t Weighted; W f , W i , W C , W o is the weight matrix, which is used to calculate the forget gate, input gate, candidate cell state and output gate respectively; b f 、bi 、b C 、b o is the bias term, which is used for the calculation of the forget gate, input gate, candidate cell state and output gate respectively; σ is the activation function; tanh is the hyperbolic tangent activation function; x t is the current input vector; h t-1 is the hidden state at the previous moment.
[0204] Reverse LSTM formula:
[0205] f′ t =σ(W′ f ·[h′ t+1 , x t ]+b′ f )
[0206] i′ t =σ(W′ i ·[h′ t+1 , x t ]+b′ i )
[0207]
[0208] o′ t =σ(W′ o ·[h′ t+1 , x t ]+b′ o )
[0209] h′ t =o′ t tanh(C′ t )
[0210] The meaning of the symbols in the reverse LSTM formula is the same as that in the forward LSTM formula. “′” represents the parameter in the reverse sequence.
[0211] Bidirectional LSTM output:
[0212] H t =[h t ; h′ t ],
[0213] Among them, H t is the bidirectional LSTM output at the current moment, which is composed of the hidden state h of the forward LSTM t and the hidden state h′ of the reverse LSTM t Splicing
[0214] Attention mechanism: Add attention mechanism to highlight the impact of data at key time steps on prediction.
[0215] Attention weights:
[0216] e t =v T tanh(W H H t +b H ),
[0217] Among them, e t is the attention score at the current moment, indicating the importance of time step ttt; v T Weight vector, used to calculate the attention score; W H is the weight matrix of the attention layer; b H is the bias of the attention layer; H t is the bidirectional LSTM output at the current moment.
[0218] Attention score:
[0219]
[0220] Among them, α t is the attention weight at the current moment, indicating the normalized importance of time step t; e t is the attention score at the current moment; T is the total length of the sequence.
[0221] Context vector:
[0222]
[0223] Where c is the context vector, which is obtained by weighted summing the attention weights and bidirectional LSTM outputs at all time steps; α t is the attention weight at the current moment. t is the bidirectional LSTM output at the current moment.
[0224] Multi-task learning: Predict hardness and thickness simultaneously and improve prediction accuracy through shared feature representation.
[0225] Hardness prediction:
[0226]
[0227] in, is the predicted hardness value; W hardness is the weight matrix of the hardness prediction layer. b hardness is the bias of the hardness prediction layer. c is the context vector.
[0228] Thickness prediction:
[0229]
[0230] in, is the predicted thickness value. Wthickness is the weight matrix of the thickness prediction layer. W thickness is the bias of the thickness prediction layer and c is the context vector.
[0231] Loss function formula:
[0232] Hardness loss:
[0233]
[0234] Among them, L hardness Loss function for hardness prediction; N is the number of samples; The predicted hardness value of the i-th sample; y hardness,i is the true hardness value of the i-th sample.
[0235] Thickness loss:
[0236]
[0237] Among them, L thickness Loss function for thickness prediction; N is the number of samples; is the predicted thickness value of the ith sample. thickness,i is the true thickness value of the i-th sample.
[0238] Total loss:
[0239] L total =λ 1 L hardness +λ 2 L thickness ,
[0240] Among them, L total is the total loss function used to optimize hardness and thickness prediction simultaneously; λ 1 and λ 2 are the weight coefficients of hardness and thickness loss, respectively, controlling their contributions to the total loss.
[0241] Step 3: Optimization algorithm application
[0242] Genetic algorithm is used to iteratively find the best combination of extrusion process parameters. In order to improve the efficiency and accuracy of extrusion process parameter optimization, the following improvements can be made to the standard genetic algorithm:
[0243] Adaptive crossover and mutation: Adaptively adjust the crossover rate and mutation rate according to fitness.
[0244] An initial population is randomly generated, each individual represents a combination of extrusion process parameters:
[0245] Population size: Set the population size to P.
[0246] Individual coding: Each individual x is encoded with a set of parameters x = (p 1 , p 2 , p 3 …p n ) indicates that p i is the extrusion process parameter.
[0247] Fitness function:
[0248] Define the fitness function f(x) to evaluate the quality of each individual, usually calculated based on hardness and thickness error:
[0249] f(x)=w 1 ·hardness error (x)+w 2 ·thickness error (x)
[0250] Among them, hardness error (x) and thickness error (x) represents the hardness and thickness error of individual x respectively. 1 and w 2 is the weight coefficient of hardness and thickness error.
[0251] Multiple selection strategy combinations: Combine roulette selection and tournament selection, taking into account both exploration and development.
[0252] Roulette wheel selection: individuals are selected with probability proportional to their fitness values.
[0253] Tournament selection: randomly select k individuals from the population and choose the one with the best fitness value as the parent.
[0254] Adaptive crossover rate is used to adjust the crossover rate according to the fitness difference of parents:
[0255]
[0256] Among them, α is the crossover rate; α max and α min are the minimum and maximum crossover rates respectively; Δf is the fitness difference between the parents. max and f min are the maximum and minimum fitness values in the current population respectively.
[0257] Single point crossover:
[0258] x new1 =α·x parent1 +(1-α)·x parent2
[0259] x new2 =(1-α)·xparent1 +α·x parent2
[0260] Among them, x new1 is a new individual 1 generated by crossover; x new2 is the new individual 2 generated by crossover; α is the crossover rate; x parent1 is the parent individual 1; x parent2 is parent individual 2;
[0261] Mutation: Adaptive mutation rate is used to adjust the mutation rate according to individual fitness.
[0262]
[0263] Among them, β is the mutation rate; β max and β min are the minimum and maximum mutation rates respectively; f(x) is the fitness value of the current individual; f max and f min are the maximum and minimum fitness values in the current population respectively.
[0264] Mutation operation:
[0265] x new =x+β·δ,
[0266] Among them, β is the mutation rate; δ is the random disturbance, which obeys the standard normal distribution; x new is the new individual after mutation; x is the original individual;
[0267] Elite retention strategy: retain the best E individuals in the previous generation and directly enter the next generation to avoid the loss of excellent genes.
[0268] Elite individuals: Select E individuals with the highest fitness values from the previous generation and directly add them to the new population.
[0269] Local search: Perform local search on the current optimal solution during the iteration process to improve the accuracy of the solution:
[0270] x′ new =x best +γ·δ,
[0271] Among them, x′ new is the new individual after local search; x best is the current optimal individual; γ is the local search step length; δ is a random disturbance, which usually obeys the standard normal distribution;
[0272] Iterative optimization: Repeat selection, crossover, mutation, elite retention, and local search operations until a stopping condition is reached.
[0273] Step 4: Real-time optimization algorithm
[0274] According to the optimization results, the parameters of the extrusion equipment are adjusted in real time to ensure that the hardness and thickness of the mattress meet the requirements.
[0275] Real-time optimization and control:
[0276] Real-time fitness calculation and parameter adjustment: Calculate the fitness value of the current extrusion process parameters in real time, and perform crossover and mutation operations to generate new parameter combinations.
[0277] According to the optimization results, the parameters of the extrusion equipment are adjusted in real time: the target extrusion force is 15kN, which is achieved by adjusting the hydraulic system. The target extrusion speed is 50mm / s, and the motor speed is controlled by the inverter.
[0278] Feedback control system:
[0279] Use PID controller to provide real-time feedback and correction of adjusted parameters:
[0280]
[0281] Among them, u(t) is the control input hydraulic valve opening and motor voltage, e(t) is the error, K p , K i , K d are the proportional, integral, and differential coefficients.
[0282] Real-time monitoring and alarm system:
[0283] Real-time monitoring system: A monitoring system is installed in the central control room to display graphs and data of squeeze force, displacement, mattress hardness and thickness in real time. Set alarm thresholds, and when parameters exceed the set range, an alarm signal is automatically issued to prompt the operator to intervene.
[0284] Alarm processing: When the system detects an abnormal situation, such as the extrusion pressure is too high or too low, or the extrusion speed is too fast or too slow, it will automatically trigger an alarm and record the abnormal data. The operator will adjust the equipment in time according to the alarm information to ensure the stability of the production process.
[0285] Data storage and analysis:
[0286] Data storage: The real-time collected data is stored in the database for subsequent analysis and optimization. The stored data includes the extrusion force, displacement, hardness, thickness and corresponding optimization parameters at each time step.
[0287] Data analysis: Regularly analyze the stored data to evaluate the stability of the production process and product quality. Use data mining and machine learning techniques to identify problems in the production process and propose improvement plans.
[0288] Numerical case analysis:
[0289] 1. Intelligent management of molding process: using improved LSTM
[0290] Data preparation: Split the data into training set 80% and test set 20%.
[0291] Model Architecture:
[0292] Bidirectional LSTM layer: Two layers of LSTM, 50 units in each layer.
[0293] Attention mechanism: Calculate the attention weight for each time step to obtain the context vector.
[0294] Output layer: predict the thickness and hardness of latex foam.
[0295] Model training results: The prediction accuracy of the trained model on the test set is:
[0296] Thickness prediction mean square error: 0.05mm
[0297] Hardness prediction mean square error: 0.2 Shore A
[0298] Real-time prediction and control: In the real-time production process, standardized pressure and temperature data are input, the LSTM model predicts the thickness and hardness, and the control system adjusts the temperature and pressure of the mold based on the prediction results.
[0299] Current Normalized Pressure: 1.5
[0300] Current normalized temperature: 0.3
[0301] Prediction results:
[0302] Thickness: 10.05mm
[0303] Hardness: 30.3Shore A
[0304] According to the prediction results, the mold temperature was adjusted to 71°C and the pressure to 5.2MPa to ensure the stability of the molding process and product quality.
[0305] 2. Intelligent optimization of extrusion process: using improved genetic algorithm
[0306] With weight w 1 =0.6, w 2 = 0.4, calculate the fitness value of each individual, and select the best individual x 2 and x 50 Perform crossover and mutation:
[0307] Crossover rate α:
[0308]
[0309] Crossover operation:
[0310] x new1 =0.9·x2+0.1·x50=0.9·(15.0,50)+0.1·(14.8,49)=(14.98,49.9)
[0311] x new2 =0.1·x2+0.9·x50=0.1·(15.0,50)+0.9·(14.8,49)=(14.82,49.1)
[0312] Mutation rate β:
[0313]
[0314] Mutation operation:
[0315] x new1 =x new1 +0.1·δ
[0316] When δ=(0.02,-0.03)
[0317] Then x new1 =(14.98,49.9)+0.1·(0.02,-0.03)=(15.0,49.87)
[0318] Real-time optimization and control:
[0319] According to the optimization results, adjust the parameters of the extrusion equipment in real time:
[0320] The current best individual: x best =(15.0, 50), adjust the extrusion pressure to 15.0 kN and the extrusion speed to 50 mm / s.
[0321] Feedback Control:
[0322] Use the PID controller to provide real-time feedback and correction to the adjusted parameters. The formula is:
[0323]
[0324] Use PID controller to provide real-time feedback and correction to the adjusted parameters to ensure control accuracy. The PID coefficient is K p =1.0, K i =0.5, K d =0.1.
[0325] Finally, the central control room monitors the production process in real time, sets alarm thresholds, and prompts operators to intervene. The stored data is analyzed regularly to evaluate the stability of the production process and product quality, and to propose improvement plans.
[0326] Example 2
[0327] This embodiment 2 provides a latex mattress production device and process thereof, which is used to further illustrate the working process or principle of the latex mattress production device and process thereof provided in the above embodiment 1, and the details are as follows:
[0328] A latex mattress production process comprises the following steps:
[0329] Sc1. First, after the motor body 401 is started, the active top gear 402 is driven to rotate by the motor body 401. Through the meshing of the active top gear 402 and the transmission side gear 403, the rotating active top gear 402 will drive the transmission side gear 403 to rotate. At the same time, during the rotation process, the transmission side gear 403 will be movably connected to the surface of the bottom support platform 405 for support through the bearing plate 404. The connection of the bearing plate 404 can improve the stability of the transmission side gear 403 during the rotation process. The rotating transmission side gear 403 will drive the mixing tank 2 to rotate. Compared with the traditional simple internal stirring, the internal latex is mixed by the rotating mixing tank 2, which is helpful to improve the uniformity of the latex mixing and the fluidity of the latex, thereby improving the overall quality and performance of the mattress;
[0330] Sc2. When the motor body 401 drives the active top gear 402 to rotate, it will drive the bottom transmission rod 501 to rotate. The rotating bottom transmission rod 501 will drive the top active gear disc 502 to rotate. Through the meshing of the top active gear disc 502 and the side transmission gear disc 503, the side transmission gear disc 503 can be driven to rotate synchronously during the rotation of the top active gear disc 502. During the rotation of the side transmission gear disc 503, the side transmission rod 504 will be driven to rotate synchronously. When the side transmission rod 504 rotates, it will drive the side movable gear disc 505 and the side connecting gear disc 506 to rotate. The direction of force rotation can be changed by using the side movable gear disc 505 and the side connecting gear disc 506, so that the rotation of the side connecting gear disc 506 can drive the bottom connecting rod 507 to rotate. When the bottom connecting rod 507 rotates, it will pass through the top limit hole provided on the surface of the top cover 3. 508 moves, and the top limiting hole 508 can be used to limit the bottom connecting rod 507, thereby reducing the deviation caused by the rotation of the bottom connecting rod 507. When the bottom connecting rod 507 rotates, it will drive the two top limiting holes 508 to rotate, and the two top limiting holes 508 will respectively drive the stirring frame 509 to rotate and stir along the inner side of the mixing tank 2. At this time, the transmission connection of the top active gear disc 502, the side transmission gear disc 503, the side transmission rod 504, the side movable gear disc 505 and the side connecting gear disc 506 can change the rotation direction of the bottom connecting rod 507 and the active top gear 402. At this time, when the mixing tank 2 rotates to mix the latex, the stirring frame 509 will rotate in the opposite direction of the mixing tank 2 to stir the latex, which can improve the effect of stirring the latex, improve the mixing property of the latex, reduce the mixing time, and improve the mixing efficiency.
[0331] Sc3. During the rotation of the bottom connecting rod 507, the top connecting frame 601 is supported by the external fixing frame 5010, and the fixing frame 5010 can be firmly connected to the top of the processing base 1 by using the side support frame 5011. At this time, after being connected to the external power supply through the battery body 603, the controller body 602 can be driven to work by the battery body 603, and the heating tube 607 can be driven to work by the controller body 602 through the connection of the transmission cable 604. The controller body 602 passes through the connecting groove 605 opened on the inner side of the top limit hole 508 and enters the heating box 606 opened on the inner side of the stirring frame 509. By driving the heating tube 607 to work, the temperature in the heating box 606 can be increased, and the increased temperature will be conducted to the latex through the heat-conducting copper tube 608, so that the temperature of the latex can be continuously increased by using the heat-conducting copper tube 608 during the stirring of the latex when the stirring frame 509 rotates, which can avoid the phenomenon of the latex becoming gradually viscous and improve the fluidity and effect of the latex.
[0332] Sc4. After the latex is stirred, liquid can be first added to the liquid storage tank 701 along the liquid inlet 702, and after being started by the water pump body 703, the liquid in the liquid storage tank 701 is driven along the top discharge pipe 705 into the cooling circulation pipe 704. The coolant will flow along the multi-layer cooling circulation pipe 704. When it flows to the bottom layer, it will enter the liquid storage tank 701 through the bottom discharge pipe 706 to form a cycle. The coolant circulating in the mixing tank 2 can quickly reduce the temperature of the mixing tank 2, so that the latex processed in the mixing tank 2 can be cooled. The circulating coolant is used to evenly flow along the inner wall of the mixing tank 2, which can improve the uniformity of cooling and improve the efficiency and effect of latex cooling.
[0333] It should be noted that the water pump body 703, controller body 602, battery body 603 and motor body 401 in the present invention are all prior art, and corresponding models can be selected according to actual needs. The internal structure and operating principle of the above parts are also common knowledge of technicians in this field, so no further elaboration will be made on this.
[0334] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A latex mattress production device, comprising a processing base (1), characterized in that: A mixing tank (2) is arranged on the top of the processing base (1), a top cover (3) is fixedly connected to the top of the mixing tank (2), a motor body (401) is fixedly connected to the surface of the processing base (1), an output shaft of the motor body (401) is movably connected to a driving top tooth (402), one side of the driving top tooth (402) is movably connected to a transmission side tooth (403), and the top of the transmission side tooth (403) is fixedly connected to the bottom of the mixing tank (2).
2. A latex mattress production device according to claim 1, characterized in that: The top of the active top tooth (402) is fixedly connected to a bottom transmission rod (501), the top of the bottom transmission rod (501) is fixedly connected to a top active toothed disc (502), one side of the top active toothed disc (502) is movably connected to a side transmission toothed disc (503), one side of the side transmission toothed disc (503) is fixedly connected to a side transmission rod (504), one end of the side transmission rod (504) is fixedly connected to a side movable toothed disc (505), one side of the side movable toothed disc (505) is movably connected to a side connecting toothed disc (506), the inner side of the side connecting toothed disc (506) is fixedly connected to a bottom connecting rod (507), the surface of the bottom connecting rod (507) is fixedly connected to a top limiting hole (508), and one side of the top limiting hole (508) is fixedly connected to a stirring frame (509).
3. A latex mattress production device according to claim 2, characterized in that: A heating box (606) is provided on the inner side of the stirring frame (509), a heating tube (607) is movably connected to the inner side of the heating box (606), a transmission cable (604) is provided on the inner side of the top limit hole (508), and a connecting groove (605) is provided on the inner side of the transmission cable (604).
4. A latex mattress production device according to claim 3, characterized in that: The top of the processing base (1) is fixedly connected to a side support frame (5011), one side of the side support frame (5011) is fixedly connected to a fixing frame (5010), a top connecting hole (5012) is opened at the top of the fixing frame (5010), and the inner side of the top connecting hole (5012) is arranged on the surface of the bottom connecting rod (507).
5. A latex mattress production device according to claim 4, characterized in that: The top of the fixing frame (5010) is fixedly connected to a top connecting frame (601), the top of the top connecting frame (601) is fixedly connected to a battery body (603), and a controller body (602) is provided on one side of the battery body (603).
6. A latex mattress production device according to claim 5, characterized in that: A cooling circulation pipe (704) is provided on the inner side of the mixing tank (2), and liquid storage tanks (701) are fixedly connected on both sides of the mixing tank (2). A liquid inlet (702) is provided on the top of the liquid storage tank (701), and a water pump body (703) is provided on one side of the liquid inlet (702). A top liquid discharge pipe (705) is fixedly connected on the other side of the liquid storage tank (701), and a bottom liquid discharge pipe (706) is provided on the bottom of the liquid storage tank (701).
7. A latex mattress production device according to claim 6, characterized in that: A heat-conducting copper tube (608) is fixedly connected to one side of the stirring frame (509), and the number of the heat-conducting copper tubes (608) is multiple.
8. The latex mattress production device according to claim 7, characterized in that: A top limiting hole (508) is provided on the surface of the top cover (3), and the inner side of the top limiting hole (508) is arranged on the surface of the bottom connecting rod (507).
9. The latex mattress production device according to claim 8, characterized in that: The bottom of the transmission side teeth (403) is movably connected to a bearing plate (404), the bottom of the bearing plate (404) is movably connected to a bottom support platform (405), and the bottom of the bottom support platform (405) is fixedly connected to the top of the processing base (1).
10. A latex mattress production process, applied to a latex mattress production device as claimed in any one of claims 1 to 9, characterized in that: The steps include: S1. Prepare latex and additives, and prepare the latex and additives according to the formula ratio; S2, adding raw materials of latex and additives into a mixing tank (2) for mixing and stirring; S3, putting the mixed raw materials into a foaming machine, stirring at high speed, adding an appropriate amount of foaming agent, and then making the latex react chemically in the mold by high temperature and pressure to form latex foam; S4. Inject the latex foam into the pre-prepared mold to automatically shape it. The mold is made according to the size and shape of the mattress to ensure the precise size of the mattress; S5. The latex mattress after forming needs to go through the extrusion process to achieve the appropriate hardness and thickness, and the latex mattress after extrusion needs to be trimmed to remove the burrs and uneven parts on the surface; S6. Clean the processed latex mattress to remove excess residue and odor; S7. Conduct quality inspection on the produced latex mattresses to ensure that the products meet the relevant standards and requirements; S8. Package the repaired latex mattress to prevent damage during transportation.
11. A latex mattress production process according to claim 10, characterized in that: According to S4 and S5, sensors are installed in the mold, and the predictive control algorithm is used to predict the molding process of latex foam in the mold in real time, and the temperature and pressure of the mold are dynamically adjusted to ensure the precise size and shape of the mattress; then, by installing force sensors and displacement sensors, the genetic algorithm is used to optimize the extrusion process parameters of pressure and extrusion speed to ensure that the latex mattress reaches the appropriate hardness and thickness, while reducing the defective rate; the specific process is: Step 1: Data collection and preprocessing Smart sensor installation Pressure sensors: installed at multiple key locations of the mold to monitor the pressure distribution throughout the molding process; Temperature sensor: installed at different locations inside and outside the mold to monitor the temperature changes of latex foam; Force sensor: installed at the key stress-bearing part of the extrusion equipment to monitor the force changes during the entire extrusion process; Displacement sensor: installed on the moving parts of the extrusion equipment to monitor the displacement changes during the extrusion process in real time; Real-time data collection: Sampling frequency: 100 times per second; Data storage: Use PLC programmable logic controller for preliminary data processing and storage, and then transmit it to the central control system for analysis; Data preprocessing: Data cleaning: remove outliers and noise that may appear during the collection process, and use a moving average filter for smoothing; Data standardization: Standardize pressure and temperature data to facilitate subsequent machine learning algorithm processing; Standardization formula: Among them, v is the original data, μ is the mean of the data, and e is the standard deviation of the data; Step 2: Establish a prediction model for the hardness and thickness of latex mattresses and use the long short-term memory network (LSTM) for training. In order to improve the accuracy of the prediction model for the hardness and thickness of latex mattresses, the following improvements can be made to the standard LSTM: Bidirectional LSTM: Bidirectional LSTM is used to consider both forward and backward information at the same time to enhance the modeling ability of time series data. It consists of two LSTM layers, one for processing the forward sequence and the other for processing the reverse sequence. The specific formula is as follows: Forward LSTM formula: f t =σ(W f ·[h t-1 ,x t ]+b f ), i t =σ(W i ·[h t-1 ,x t ]+b i ), the t =σ(W o ·[h t-1 ,x t ]+b o ) h t =o t ·tanh(C t ) Among them, f t is the activation vector of the forget gate, which determines the cell state C t-1 The degree to which it needs to be forgotten; t is the activation vector of the input gate, which determines the new information to what extent the cell state is updated; is the candidate cell state, containing new candidate information; C t is the current cell state, which is obtained by weighted sum of the cell state at the previous moment and the current candidate state; o t is the activation vector of the output gate, which determines the cell state C t To what extent does it affect the hidden state h t ;h t is the current hidden state, through the cell state C t Weighted; W f , W i , W C , W o is the weight matrix, which is used to calculate the forget gate, input gate, candidate cell state and output gate respectively; b f , b i , b C , b o is the bias term, which is used for the calculation of the forget gate, input gate, candidate cell state and output gate respectively; σ is the activation function; tanh is the hyperbolic tangent activation function; x t is the current input vector; h t-1 is the hidden state at the previous moment; Reverse LSTM formula: f′ t =σ(W′ f ·[h′ t+1 ,x t ]+b′ f ) i′ t =σ(W′ i ·[h′ t+1 ,x t ]+b′ i ) oh t =σ(W′ o ·[h′ t+1 ,x t ]+b′ o ) h t =o' t ·tanh(C′ t ) The meaning of the symbols in the reverse LSTM formula is the same as that in the forward LSTM formula, and "′" represents the parameter in the reverse sequence; Bidirectional LSTM output: H t =[h t ;h′ t ], Among them, H t is the bidirectional LSTM output at the current moment, which is composed of the hidden state h of the forward LSTM t and the hidden state h′ of the reverse LSTM t Splicing obtained; Attention mechanism: Add attention mechanism to highlight the impact of data at key time steps on prediction; Attention weights: e t =v T tanh(W H H t +b H ), Among them, e t is the attention score at the current moment, indicating the importance of time step t; v T Weight vector, used to calculate the attention score; W H is the weight matrix of the attention layer; b H is the bias of the attention layer; H t is the bidirectional LSTM output at the current moment; Attention score: Among them, α t is the attention weight at the current moment, indicating the normalized importance of time t; e t is the attention score at the current moment; T is the total length of the sequence; Context vector: Where c is the context vector, which is obtained by weighted summing the attention weights and bidirectional LSTM outputs at all time steps; α t is the attention weight at the current moment; H t is the bidirectional LSTM output at the current moment; Multi-task learning: predict hardness and thickness simultaneously, and improve prediction accuracy through shared feature representation; Hardness prediction: in, is the predicted hardness value; W hardness is the weight matrix of the hardness prediction layer; b hardness is the bias of the hardness prediction layer; c is the context vector; Thickness prediction: in, is the predicted thickness value; W thickness is the weight matrix of the thickness prediction layer; W thickness is the bias of the thickness prediction layer; Loss function formula: Hardness loss: Among them, L hardness Loss function for hardness prediction; N is the number of samples; The predicted hardness value of the i-th sample; y hardness,i is the true hardness value of the i-th sample; Thickness loss: Among them, L thickness Loss function for thickness prediction; N is the number of samples; is the predicted thickness value of the i-th sample; y thickness,i is the true thickness value of the i-th sample; Total loss: L total =λ1L hardness +λ2L thickness , Among them, L tota1 is the total loss function, which is used to optimize hardness and thickness prediction simultaneously; λ1 and λ2 are the weight coefficients of hardness and thickness loss, respectively, controlling their contributions to the total loss; Step 3: Optimization algorithm application Genetic algorithms are used to iteratively find the best combination of extrusion process parameters. In order to improve the efficiency and accuracy of extrusion process parameter optimization, the following improvements can be made to the standard genetic algorithm: Adaptive crossover and mutation: adaptively adjust the crossover rate and mutation rate according to fitness; An initial population is randomly generated, each individual represents a combination of extrusion process parameters: Population size: Set the population size to P; Individual coding: Each individual x is encoded with a set of parameters x = (p1, p2, p3...p n ) indicates that p i is the extrusion process parameter; Fitness function: Define the fitness function f(x) to evaluate the quality of each individual, usually calculated based on hardness and thickness error: f(x)=w1·hardness error (x)+w2·thickness error (x) Among them, hardness error (x) and thickness error (x) represents the hardness and thickness errors of individual x, respectively; w1 and w2 are the weight coefficients of hardness and thickness errors; Multiple selection strategy combinations: combining roulette selection and tournament selection, taking into account both exploration and development; Roulette wheel selection: select individuals based on the proportional probability of their fitness values; Tournament selection: randomly select k individuals from the population and choose the individual with the best fitness value as the parent; Adaptive crossover rate is used to adjust the crossover rate according to the fitness difference of parents: Among them, α is the crossover rate; α max and α min are the minimum and maximum crossover rates respectively; Δf is the fitness difference between the parents; f max and f min are the maximum and minimum fitness values in the current population respectively; Single point crossover: x new1 =α·x parent1 +(1-a)·x parent2 x new2 =(1-a)·x parent1 +a·x parent2 Among them, x new1 is a new individual 1 generated by crossover; x new2 is the new individual 2 generated by crossover; α is the crossover rate; x parent1 is the parent individual 1; x parent2 is parent individual 2; Mutation: Adopt adaptive mutation rate and adjust the mutation rate according to individual fitness; Among them, β is the mutation rate; β max and β min are the minimum and maximum mutation rates respectively; f(x) is the fitness value of the current individual; f max and f min are the maximum and minimum fitness values in the current population respectively; Mutation operation: x new =x+β·δ, Among them, β is the mutation rate; δ is the random disturbance, which obeys the standard normal distribution; x new is the new individual after mutation; x is the original individual; Elite retention strategy: retain the best E individuals in the previous generation and directly enter the next generation to avoid the loss of good genes; Elite individuals: select E individuals with the highest fitness values from the previous generation and directly add them to the new population; Local search: Perform local search on the current optimal solution during the iteration process to improve the accuracy of the solution: x′ new =x best +g·d, Among them, x′ new is the new individual after local search; x best is the current optimal individual; γ is the local search step length; δ is a random disturbance, which usually obeys the standard normal distribution; Iterative optimization: Repeat selection, crossover, mutation, elite retention, and local search operations until a stopping condition is reached; Step 4: Real-time optimization algorithm According to the optimization results, the parameters of the extrusion equipment are adjusted in real time to ensure that the hardness and thickness of the mattress meet the requirements; Real-time optimization and control: Real-time fitness calculation and parameter adjustment: Real-time calculation of the fitness value of the current extrusion process parameters, and crossover and mutation operations to generate new parameter combinations; According to the optimization results, the parameters of the extrusion equipment are adjusted in real time: the target extrusion force is 15kN, which is achieved by adjusting the hydraulic system; the target extrusion speed is 50mm / s, and the motor speed is controlled by the frequency converter; Feedback control system: Use PID controller to provide real-time feedback and correction of adjusted parameters: Among them, u(t) is the control input hydraulic valve opening and motor voltage, e(t) is the error, K p , K i , K d are proportional, integral and differential coefficients; Real-time monitoring and alarm system: Real-time monitoring system: A monitoring system is installed in the central control room to display graphs and data of squeeze force, displacement, mattress hardness and thickness in real time; alarm thresholds are set, and when the parameters exceed the set range, an alarm signal is automatically issued to prompt the operator to intervene; Alarm processing: When the system detects abnormal conditions, such as the extrusion pressure is too high or too low, or the extrusion speed is too fast or too slow, the alarm is automatically triggered and the abnormal data is recorded; the operator adjusts the equipment in time according to the alarm information to ensure the stability of the production process; Data Storage and Analysis: Data storage: Store the real-time collected data in the database for subsequent analysis and optimization; The stored data include extrusion force, displacement, hardness, thickness and corresponding optimization parameters at each time step; Data analysis: Regularly analyze stored data to evaluate the stability of the production process and product quality; use data mining and machine learning techniques to identify problems in the production process and propose improvement plans.
Citation Information
Patent Citations
Latex product for mattresses or pillows and preparation method thereof
CN103980570A
Preparation method and application method of high-performance asphalt mixture
CN117901294A
Terylene three-dimensional crimped organosilicon emulsion preparation equipment and use method thereof
CN117983090A
Production, processing and diluting equipment for waterproof coating raw materials
CN219897863U
Temp. profile sensor for extruder nozzle or die - is shaped to suit specific task, carried in low conductivity material and can be used for manual or automatic optimisation of temp. control
DE4129979A1
Cited By
Homogenization cooling control method and system for aluminum alloy extruded part
CN120292817A
