Composite building energy-saving thermal insulation board and preparation method thereof
Through the combined structure of cement fiberglass felt, rock wool board and polyurethane board, combined with intelligent control algorithm and deep neural network optimization process parameters, the problem of insufficient bonding performance of composite insulation boards is solved, and efficient and low-cost building energy-saving insulation is achieved Thermal insulation effect.
Patent Information
- Application Number
- CN202411174989.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-12-03
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing composite insulation boards have insufficient bonding performance and complex production processes during use, resulting in high production costs and difficult to meet the building's thermal insulation and fire safety requirements.
Using a combined structure of cement glass fiber felt, rock wool board, polyurethane board and polyurethane adhesive, polyurethane black and white material are poured through a high-pressure machine to form polyurethane board, and castor oil and catalyst are added to the polyurethane adhesive, using intelligent control algorithms and depth Neural network optimizes material ratio and process parameters to improve bonding strength and thermal insulation performance.
It significantly improves the fire resistance, bonding and thermal insulation properties of composite panels, simplifies production processes, reduces costs, is suitable for large-scale applications, and extends service life.
Smart Images

Figure CN119062017A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of building materials, in particular to a composite building energy-saving thermal insulation board and a preparation method thereof. Background Art
[0002] In recent years, my country's building energy-saving wall insulation market has developed rapidly, and insulation system products have emerged in an endless stream. However, in the face of sudden fire accidents, many insulation materials are very fragile. How to ensure the fireproof performance of building materials and be able to prevent fire in the event of a sudden fire has become a very important indicator for measuring whether a building is qualified.
[0003] A Chinese patent with publication number CN102877560A discloses a composite insulation board, which includes a first fire-resistant insulation layer, a rigid foam polyurethane core material and a second fire-resistant insulation layer, wherein the first fire-resistant insulation layer is bonded to one side of the rigid foam polyurethane core material and the second fire-resistant insulation layer is bonded to the other side; the invention has a simple, compact and reasonable structure; is easy and quick to use; has good fire resistance; has good insulation and energy-saving effects; has strong tensile strength, good toughness and strength, and improves service life. The Chinese patent with publication number CN106567468A discloses a composite insulation board, which is composed of a protective board, a modified molded polystyrene foam board, a molded polystyrene foam board or an extruded polystyrene foam board. The modified molded polystyrene foam board is prepared by adjusting the superheated water steam pressure in the molding machine. After the fireproof material is poured into the cavity of the modified foam board and a fireproof layer is formed, the board is pressed or bonded with the protective board, the molded polystyrene foam board or the extruded polystyrene foam board to form a composite insulation board, so as to achieve an organic combination of Class A fireproofing, safety and fire protection requirements, and excellent insulation effect. The composite insulation board of the present invention can meet the insulation and fire safety requirements of buildings in different latitudes, and can further reduce the thermal conductivity of the insulation layer, reduce the design thickness of the insulation layer, increase the use space of the house, and further improve the cost performance and social benefits. However, during the use of the composite board prepared by the above technical solution, the bonding performance between the layers needs to be improved, and the manufacturing process is relatively complicated, which increases the production cost. Therefore, the present invention provides a composite building energy-saving thermal insulation board and a preparation method thereof to solve the above problems. Summary of the invention
[0004] In view of the deficiencies in the prior art, the present invention provides a composite building energy-saving thermal insulation board and a preparation method thereof, which solves the problems mentioned in the above background technology.
[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a composite insulation board, which includes a cement fiberglass felt, an insulation layer, a rock wool board, a polyurethane board and a polyurethane adhesive;
[0006] The surface of the cement glass fiber felt is coated with a polyurethane adhesive, a rock wool board is pasted on the polyurethane adhesive, an upper polyurethane board is arranged on the rock wool board by a high pressure machine, and an insulation layer is covered on the surface of the polyurethane board.
[0007] Preferably, black polyurethane material and white polyurethane material are poured on the rock wool board and foamed to form a polyurethane board, wherein the mass ratio of the black polyurethane material to the white polyurethane material is (1.2-1.6):1.
[0008] Preferably, the thermal insulation layer is prepared by adding SiO2 aerogel to anhydrous ethanol, stirring, and then adding KH550 and Tween 80 until they are evenly mixed to obtain the thermal insulation layer.
[0009] A method for preparing a composite building energy-saving thermal insulation board, specifically comprising the following preparation steps:
[0010] Step 1: Coat the 2.0 mm thick cement fiberglass felt with polyurethane adhesive, and stick the 50 mm thick rock wool board on the polyurethane adhesive;
[0011] Step 2: Then put it into a laminator, and at the entrance of the laminator, use a high-pressure machine to pour polyurethane black material and white material on the rock wool board, and foam it to form a 150mm thick polyurethane board;
[0012] Step 3: Under normal temperature and pressure, place the material obtained in step 2 in an insulation layer for infiltration treatment, then place it in a fume hood for volatilization at room temperature for 24 to 30 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
[0013] Preferably, the preparation method of the polyurethane adhesive is:
[0014] S1. Place 15-20 parts of polyether polyol in a drying oven at 120-140°C for 2-3 hours, then take it out and cool it down to 70-80°C;
[0015] S2, mixing the dried polyether polyol with 10-15 parts of isophorone diisocyanate, placing in a dryer, heating to 80-90° C., and reacting for 1.5-2.5 hours;
[0016] S3, adding 0.03-0.05 parts of catalyst, reacting for 0.5-1.5 hours, and then adding 0.1-0.2 parts of triethylene glycol;
[0017] S4, mixing the product in S3 with 3-5 parts of castor oil, and placing the mixture in a drying oven at 120-130° C. for dehydration and drying;
[0018] S5, pour in 0.4-1 part of chain extender, stir for 8-10 minutes until it is evenly mixed, add auxiliary agent, continue stirring, then place in a degassed environment and let stand for 5-10 minutes, and introduce nitrogen again;
[0019] S6. Place the product in S5 in a drying oven at 120° C. and cure for 3 to 4 hours to obtain a polyurethane adhesive.
[0020] Preferably, the catalyst in S3 is a mixture of stannous octoate and dibutyltin dilaurate.
[0021] Preferably, the auxiliary agent in S5 is dimethyl carbonate and diphenylmethane 4,4-diisocyanate, and the two are added sequentially.
[0022] Preferably, the lamination temperature in step 2 is 55-65° C., and the conveying speed during the lamination process is 4.5-5.5 m / min.
[0023] Preferably, in steps 1 to 3, the production process is optimized using an intelligent control algorithm, which specifically includes the following steps:
[0024] Step 1: Define optimization objectives and variables:
[0025] Comprehensive data collection through sensors:
[0026] Temperature sensor: monitors temperature changes during the production process;
[0027] Pressure sensor: measures the pressure inside the production equipment;
[0028] Humidity sensor: monitors ambient humidity;
[0029] Thickness sensor: measures the thickness of each layer of the insulation board;
[0030] Flow sensor: monitors the flow ratio of black and white polyurethane materials
[0031] Thermal insulation performance test: Use heat conduction equipment to measure the thermal insulation effect;
[0032] Bond strength test: Measure the bond strength using a tensile testing machine;
[0033] Durability testing: Conduct accelerated aging tests to measure the durability of the product;
[0034] Raw material consumption: record the consumption of polyurethane black material, white material and other raw materials;
[0035] Optimization goals: maximize thermal insulation performance P1, maximize bonding strength P2, maximize durability P3, minimize cost C, and maximize production efficiency E;
[0036] Variable: material ratio x1, x2, ..., x n ; Process parameters y1, y2, ..., y m ; Set the particle swarm size N, the initial position and velocity of each particle;
[0037] x i =(x i1 , x i2 , ..., x in ), v i =(v i1 , v i2 , ..., v in ), i = 1, 2, ..., N
[0038] Among them, x in is the position of the ith particle in the nth dimension; v in is the current velocity of the ith particle in the nth dimension; calculate the fitness function value f(x i );
[0039] f(x i )=w1P1+w2P2+w3P3-w4C+w5E
[0040] Among them, w1, w2, w3, w4, w5 are the weights of each target;
[0041] Update individual optimal and global optimal positions;
[0042] p i =min(f(x i )), g = min(f(p i ))
[0043] Among them, p i The individual optimal position of the i-th particle, g the global optimal position, that is, the optimal position among all particles, f(p i ) i The fitness function value of each particle, min(·) takes the minimum value function;
[0044] Update speed and location:
[0045] v ij (t+1)=wv ij (t)+c1r1[p ij -x ij (t)]+c2r2[g j -x ij (t)]
[0046] x ij (t+1)=x ij (t)+vij (t+1)
[0047] Among them, w is the inertia weight, c1, c2 are learning factors, r1, r2 are random numbers, v ij (t+1) and x ij (t+1) is the velocity and position of the i-th particle in the j-th dimension at the next time step, v ij (t) is the velocity of the i-th particle in the j-th dimension at the current step, p ij is the individual optimal velocity of the ith particle in the jth dimension, x ij (t) is the current position of the ith particle in the jth dimension, g j is the global optimal position of all particles in the jth dimension, t represents the current time step, and t+1 represents the next step after the current step;
[0048] Step 2: Use the deep neural network DNN to predict the performance indicators of different combinations and use them as the objective function of the particle swarm PSO:
[0049] Input layer: 2 neurons, corresponding to material ratio and process parameters;
[0050] Hidden layer: 128 hidden layers are set, each hidden layer contains 8 neurons, and the activation function is ReLU;
[0051] Output layer: 2 neurons, corresponding to performance indicators;
[0052] Step 3: Use deep deterministic policy gradient (DDPG) to dynamically adjust parameters in the production process, and use PID controller for fine-tuning to ensure the stability of the production process.
[0053] State s: material combination and process parameters;
[0054] Action a: Adjust material ratio and process parameters;
[0055] Reward r: Define the reward function based on product performance and cost;
[0056] Select an action:
[0057]
[0058] in, is the exploration noise at time step t, a t is the action at time step t, s t is the state at time step t, θ μ represents the weight of the target network, μ(s t |θ μ ) indicates that the policy network (Actor network) is based on the current state s tand its weight θ μ Generate actions; perform actions and observe rewards and new states;
[0059] r t =R(s t , a t ), s t+1 =f(s t , a t )
[0060] Among them, r t Indicates that action a is executed at time step t t Reward obtained after t , a t ) represents the reward function, according to the current state s t and action a t Calculate Rewards;s t+1 Indicates that action a is executed at time step t t The new state after f(s t , a t ) represents the state transition function, which describes the transition from state s t Transfer to new state s t+1 process;
[0061] Storage experience:
[0062] D←D∪(s t , a t , r t ,s t+1 )
[0063] (s i , a i , r i ,s i+1 )~D
[0064] Where D is the experience replay pool, which stores state transition samples; (s t , a t , r t ,s t+1 ) is the empirical sample observed at time step t; (s i , a i , r i ,s i+1 ) represents an experience sample sampled from the experience replay pool;
[0065] Update the Critic network:
[0066] y i =r i +γQ′(s i+1 ,μ′(s i+1 |θ μ′)|θ Q′ )
[0067] Among them, y i Represents the target value, used to update the Critic network, r i represents the reward; γ represents the discount factor, which measures the importance of future rewards; Q′(s i+1 ,μ′(s i+1 |θ μ′ )|θ Q′ ) represents the output of the target Critic network, and evaluates the new state s i+1 The action μ′(s) generated by the execution policy network is i+1 |θ μ′ ) value; Q′ represents the target Critic network, which is a delayed update version of the Critic network used for stable training; μ′(s i+1 |θ μ′ ) The target Actor network generates s according to the state i+1 The target Actor network is a delayed updated version of the Actor network, θ μ′ Represents the parameters of the target Actor network; θ Q′ Represents the parameters of the target Critic network;
[0068]
[0069] Among them, L represents the loss function of the Critic network, Q(s i , a i |θ Q ) represents the output of the Critic network, evaluating the state s i Next, perform action a i The value of N represents the batch size, θ Q It is the Critic network;
[0070]
[0071] in, Represents the gradient of the policy network, which is used to update the Actor network parameters θ μ ; Represents the gradient of the Critic network relative to action a, indicating that in state s i Next, take action a = μ(s i ) is the gradient of the value function when Q and θ μ are the parameters of the Critic network and the Actor network respectively; μ(s i ) indicates that the policy network is based on the state s i The actions generated; Actor network with respect to parameter θ μ The gradient of
[0072] Soft update target network:
[0073] θ′←τθ+(1-τ)θ′
[0074] Among them, θ represents the parameters of the current network, θ′ represents the parameters of the target network, and τ represents the soft update coefficient;
[0075] Step 4: Adjust the parameters of the production equipment through the PID controller;
[0076]
[0077]
[0078] e T (t) = r T (t)-T(t)
[0079] e P (t) = r P (t)-P(t)
[0080] Among them, e T (t) and e P (t) represents the temperature error and pressure error at time t, r T (t) and r P (t) target temperature and target pressure at time t, T(t) and P(t) actual temperature and actual pressure at time t, u T (t) and u P (t) Control output at time t, used to adjust temperature and pressure, K p Represents proportional gain, K i Indicates the integral gain, K d represents the differential gain, and Represents the error change rate of temperature and pressure at time t.
[0081] The present invention provides a composite building energy-saving thermal insulation board and a preparation method thereof. Compared with the prior art, it has the following beneficial effects:
[0082] (1) The composite building energy-saving thermal insulation board reduces the combustion performance of polyurethane through the composite setting of rock wool board and polyurethane board, thereby achieving the purpose of fire prevention and improving the fire resistance of the composite board; through the design of the insulation layer, the waterproof performance and thermal insulation performance of the prepared composite board are effectively improved, thereby improving the mechanical properties of the composite board; and the preparation process of the technical solution is simple, the cost is low, and the equipment is not complicated, which is suitable for large-scale application production.
[0083] (2) The composite building energy-saving thermal insulation board and its preparation method can increase the peeling force of polyurethane by using castor oil, thereby improving the bonding performance of polyurethane; at the same time, under the synergistic effect of the catalyst, the bonding performance of the polyurethane adhesive can be significantly improved, so that the bonding between the layers of the board is better.
[0084] (3) By optimizing the ratio of black and white polyurethane materials and the pouring temperature, the intelligent control algorithm can accurately adjust the material ratio and process parameters to ensure the density and uniformity of the polyurethane layer, thereby significantly improving the thermal insulation performance; using DNN prediction and DDPG dynamic adjustment, the preparation process of the polyurethane adhesive and the ratio of the catalyst are optimized to make the bonding between the layers of materials stronger; by finely adjusting the combination of each layer of material and the process parameters, the stability and anti-aging performance of the composite insulation board in different environments are ensured, thereby extending the service life; through the multi-objective particle swarm optimization algorithm (PSO), a trade-off is made between materials and process parameters to find the best balance between cost and performance, effectively reducing raw material and production costs; using PID controllers to monitor and adjust the temperature and pressure in the production process in real time to ensure the stability and efficiency of the production process, thereby improving production efficiency; the intelligent control algorithm can monitor and adjust various parameters in the production process in real time to ensure that the performance and quality of each batch of products are consistent, reducing fluctuations and unqualified rates in production. BRIEF DESCRIPTION OF THE DRAWINGS
[0085] Figure 1 A schematic diagram of the composite plate structure provided by the present invention;
[0086] Figure 2 A comparison chart of water absorption test provided by the present invention;
[0087] Figure 3 This is a flow chart of the intelligent control algorithm provided by the present invention.
[0088] In the figure: 1-cement fiberglass felt, 2-insulation layer, 3-rock wool board, 4-polyurethane board, 5-polyurethane adhesive. DETAILED DESCRIPTION
[0089] 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.
[0090] In the present technical solution, the preparation method of the thermal insulation layer 2 is as follows: add 5 g of SiO2 aerogel to 50 mmL of anhydrous ethanol, stir, and then add 10 mL of KH550 and 3 g of Tween 80 until they are evenly mixed to obtain the thermal insulation layer 2.
[0091] In the technical solution, polyether 1000 or polyether 2000 is used; the chain extender is ethylenediamine.
[0092] Example 1
[0093] A method for preparing a composite building energy-saving thermal insulation board, specifically comprising the following preparation steps:
[0094] Step 1: Coat the 2.0 mm thick cement glass fiber felt 1 with polyurethane adhesive 5, and adhere the 50 mm thick rock wool board 3 on the polyurethane adhesive 5;
[0095] Step 2: Then put it into a laminator, the lamination temperature is 55°C, the conveying speed is 4.5m / min, and at the entrance of the laminator, a high pressure machine is used to pour polyurethane black material and white material on the rock wool board 3, wherein the polyurethane black material and white material are in a mass ratio of 1.2:1, and foamed to form a 150mm thick polyurethane board 4;
[0096] Step 3: Under normal temperature and pressure, place the material obtained in step 2 in the insulation layer 2 for infiltration treatment, then place it in a fume hood for volatilization at room temperature for 24 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
[0097] In this technical solution, the preparation method of the polyurethane adhesive is:
[0098] S1. Place 15 parts of polyether polyol in a drying oven at 120°C for 2 hours, then take it out and cool it to 70°C;
[0099] S2, mixing the dried polyether polyol with 15 parts of isophorone diisocyanate, placing in a dryer, heating to 80° C., reacting for 1.5 hours,;
[0100] S3, adding 0.05 parts of stannous octoate and dibutyltin dilaurate, reacting for 0.5 hours, and then adding 0.2 parts of triethylene glycol;
[0101] S4, mixing the product in S3 with 5 parts of castor oil, and placing in a drying oven at 120° C. for dehydration and drying;
[0102] S5, pour 1 part of chain extender, stir for 8 minutes until it is evenly mixed, add 0.1 parts of dimethyl carbonate and diphenylmethane 4,4-diisocyanate in sequence, continue stirring, then place in a degassed environment and stand for 5 minutes, and introduce nitrogen again;
[0103] S6. Place the product in S5 in a drying oven at 120° C. and cure for 3 hours to obtain a polyurethane adhesive.
[0104] Example 2
[0105] A method for preparing a composite building energy-saving thermal insulation board, specifically comprising the following preparation steps:
[0106] Step 1: Coat the 2.0 mm thick cement glass fiber felt 1 with polyurethane adhesive 5, and adhere the 50 mm thick rock wool board 3 on the polyurethane adhesive 5;
[0107] Step 2: Then put it into a laminator, the lamination temperature is 60°C, the conveying speed is 5m / min, and at the entrance of the laminator, a high pressure machine is used to pour polyurethane black material and white material on the rock wool board 3, wherein the polyurethane black material and white material are in a mass ratio of 1.4:1, and foamed to form a 150mm thick polyurethane board 4;
[0108] Step 3: Under normal temperature and pressure, place the material obtained in step 2 in the insulation layer 2 for infiltration treatment, then place it in a fume hood for volatilization at room temperature for 27 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
[0109] In this technical solution, the preparation method of the polyurethane adhesive is:
[0110] S1. Place 18 parts of polyether polyol in a drying oven at 130°C for 2.5 hours, then take it out and cool it to 75°C;
[0111] S2, mixing the dried polyether polyol with 12 parts of isophorone diisocyanate, placing in a dryer, heating to 85° C., reacting for 2 hours,;
[0112] S3, adding 0.04 parts of stannous octoate and dibutyltin dilaurate, reacting for 1 hour, and then adding 0.1 parts of triethylene glycol;
[0113] S4, mixing the product in S3 with 4 parts of castor oil, and placing in a drying oven at 125° C. for dehydration and drying;
[0114] S5, pour in 0.7 parts of chain extender, stir for 9 minutes until it is evenly mixed, then add 0.08 parts of dimethyl carbonate and diphenylmethane 4,4-diisocyanate in sequence, continue stirring, then place in a degassed environment and let stand for 7 minutes, and introduce nitrogen again;
[0115] S6. Place the product in S5 in a drying oven at 120° C. and cure for 3.5 hours to obtain a polyurethane adhesive.
[0116] Example 3
[0117] A method for preparing a composite building energy-saving thermal insulation board, specifically comprising the following preparation steps:
[0118] Step 1: Coat the 2.0 mm thick cement glass fiber felt 1 with polyurethane adhesive 5, and adhere the 50 mm thick rock wool board 3 on the polyurethane adhesive 5;
[0119] Step 2: Then put it into a laminator, the lamination temperature is 65°C, the conveying speed is 5.5m / min, and at the entrance of the laminator, a high pressure machine is used to pour polyurethane black material and white material on the rock wool board 3, wherein the polyurethane black material and white material are in a mass ratio of 1.6:1, and foamed to form a 150mm thick polyurethane board 4;
[0120] Step 3: Under normal temperature and pressure, place the material obtained in step 2 in the insulation layer 2 for infiltration treatment, then place it in a fume hood for volatilization at room temperature for 30 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
[0121] In this technical solution, the preparation method of the polyurethane adhesive is:
[0122] S1. Place 20 parts of polyether polyol in a drying oven at 140°C for 3 hours, then take it out and cool it down to 80°C;
[0123] S2, mixing the dried polyether polyol with 10 parts of isophorone diisocyanate, placing in a dryer, heating to 90° C., reacting for 2.5 hours,;
[0124] S3, adding 0.03 parts of stannous octoate and dibutyltin dilaurate, reacting for 1.5 hours, and then adding 0.1 parts of triethylene glycol;
[0125] S4, mixing the product in S3 with 3 parts of castor oil, and placing the mixture in a drying oven at 130° C. for dehydration and drying;
[0126] S5, pour in 0.4 parts of chain extender, stir for 10 minutes until it is evenly mixed, add 0.05 parts of dimethyl carbonate and diphenylmethane 4,4-diisocyanate in sequence, continue stirring, then place in a degassed environment and stand for 10 minutes, and introduce nitrogen again;
[0127] S6. Place the product in S5 in a drying oven at 120° C. and cure for 4 hours to obtain a polyurethane adhesive.
[0128] In the technical scheme, a method for preparing a composite building energy-saving thermal insulation board uses an intelligent control algorithm to optimize the production process, specifically comprising the following steps:
[0129] Step 1: Define optimization objectives and variables:
[0130] Comprehensive data collection through sensors:
[0131] Temperature sensor: monitors temperature changes during the production process;
[0132] Pressure sensor: measures the pressure inside the production equipment;
[0133] Humidity sensor: monitors ambient humidity;
[0134] Thickness sensor: measures the thickness of each layer of the insulation board;
[0135] Flow sensor: monitors the flow ratio of black and white polyurethane materials
[0136] Thermal insulation performance test: Use heat conduction equipment to measure the thermal insulation effect;
[0137] Bond strength test: Measure the bond strength using a tensile testing machine;
[0138] Durability testing: Conduct accelerated aging tests to measure the durability of the product;
[0139] Raw material consumption: record the consumption of polyurethane black material, white material and other raw materials;
[0140] Optimization goals: maximize thermal insulation performance P1, maximize bonding strength P2, maximize durability P3, minimize cost C, and maximize production efficiency E;
[0141] Variable: material ratio x1, x2, ..., x n ; Process parameters y1, y2, ..., y m ; Set the particle swarm size N, the initial position and velocity of each particle;
[0142] x i =(x i1 , x i2 , ..., x in ), v i =(v i1 , v i2 , ..., v in ), i = 1, 2, ..., N
[0143] Among them, x in is the position of the ith particle in the nth dimension; v in is the current velocity of the ith particle in the nth dimension; calculate the fitness function value f(x i );
[0144] f(xi )=w1P1+w2P2+w3P3-w4C+w5E
[0145] Among them, w1, w2, w3, w4, w5 are the weights of each target;
[0146] Update individual optimal and global optimal positions;
[0147] p i =min(f(x i )), g = min(f(p i ))
[0148] Among them, p i The individual optimal position of the i-th particle, g the global optimal position, that is, the optimal position among all particles, f(p i ) i The fitness function value of each particle, min(·) takes the minimum value function;
[0149] Update speed and location:
[0150] v ij (t+1)=wv ij (t)+c1r1[p ij -x ij (t)]+c2r2[g j -x ij (t)]
[0151] x ij (t+1)=x ij (t)+v ij (t+1)
[0152] Among them, w is the inertia weight, c1, c2 are learning factors, r1, r2 are random numbers, v ij (t+1) and x ij (t+1) is the velocity and position of the i-th particle in the j-th dimension at the next time step, v ij (t) is the velocity of the i-th particle in the j-th dimension at the current step, p ij is the individual optimal velocity of the ith particle in the jth dimension, x ij (t) is the current position of the ith particle in the jth dimension, g j is the global optimal position of all particles in the jth dimension, t represents the current time step, and t+1 represents the next step after the current step;
[0153] Step 2: Use deep neural network DNN to predict the performance indicators of different combinations and use it as the objective function of PSO:
[0154] Input layer: 2 neurons, corresponding to material ratio and process parameters;
[0155] Hidden layer: 128 hidden layers are set, each hidden layer contains 8 neurons, and the activation function is ReLU;
[0156] Output layer: 2 neurons, corresponding to performance indicators;
[0157] Step 3: Use deep deterministic policy gradient (DDPG) to dynamically adjust parameters in the production process, and use PID controller for fine-tuning to ensure the stability of the production process.
[0158] State s: material combination and process parameters;
[0159] Action a: Adjust material ratio and process parameters;
[0160] Reward r: Define the reward function based on product performance and cost;
[0161] Select an action:
[0162]
[0163] in, is the exploration noise at time step t, a t is the action at time step t, s t is the state at time step t, θ μ represents the weight of the target network, μ(s t |θ μ ) indicates that the policy network (Actor network) is based on the current state s t and its weight θ μ Generate actions; perform actions and observe rewards and new states;
[0164] r t =R(s t , a t ), s t+1 =f(s t , a t )
[0165] Among them, r t Indicates that action a is executed at time step t t Reward obtained after t , a t ) represents the reward function, according to the current state s t and action a t Calculate Rewards;s t+1 Indicates that action a is executed at time step t t The new state after f(s t , a t) represents the state transition function, which describes the transition from state s t Transfer to new state s t+1 process;
[0166] Storage experience:
[0167] D←D∪(s t , a t , r t ,s t+1 )
[0168] (s i , a i , r i ,s i+1 )~D
[0169] Where D is the experience replay pool, which stores state transition samples; (s t , a t , r t ,s t+1 ) is the empirical sample observed at time step t; (s i , a i , r i ,s i+1 ) represents an experience sample sampled from the experience replay pool;
[0170] Update the Critic network:
[0171] y i =r i +γQ′(s i+1 ,μ′(s i+1 |θ μ′ )|θ Q′ )
[0172] Among them, y i Represents the target value, used to update the Critic network, r i represents the reward; γ represents the discount factor, which measures the importance of future rewards; Q′(s i+1 ,μ′(s i+1 |θ μ′ )|θ Q′ ) represents the output of the target Critic network, and evaluates the new state s i+1 The action μ′(s) generated by the execution policy network is i+1 |θ μ′ ) value; Q′ represents the target Critic network, which is a delayed update version of the Critic network used for stable training; μ′(s i+1 |θ μ′ ) The target Actor network generates s according to the state i+1The target Actor network is a delayed updated version of the Actor network, θ μ′ Represents the parameters of the target Actor network; θ Q′ Represents the parameters of the target Critic network;
[0173]
[0174] Among them, L represents the loss function of the Critic network, Q(s i , a i |θ Q ) represents the output of the Critic network, evaluating the state s i Next, perform action a i The value of N represents the batch size, θ Q It is the Critic network;
[0175]
[0176] in, Represents the gradient of the policy network, which is used to update the Actor network parameters θ μ ; Represents the gradient of the Critic network relative to action a, indicating that in state s i Next, take action a = μ(s i ) is the gradient of the value function when Q and θ μ are the parameters of the Critic network and the Actor network respectively; μ(s i ) indicates that the policy network is based on the state s i The actions generated; Actor network with respect to parameter θ μ The gradient of
[0177] Soft update target network:
[0178] θ′←τθ+(1-τ)θ′
[0179] Among them, θ represents the parameters of the current network, θ′ represents the parameters of the target network, and τ represents the soft update coefficient;
[0180] Step 4: Adjust the parameters of the production equipment through the PID controller;
[0181]
[0182]
[0183] e T (t) = r T (t)-T(t)
[0184] e P (t) = r P (t)-P(t)
[0185] Among them, e T (t) and e P (t) represents the temperature error and pressure error at time t, r T (t) and r P (t) target temperature and target pressure at time t, T(t) and P(t) actual temperature and actual pressure at time t, u T (t) and u P (t) Control output at time t, used to adjust temperature and pressure, K p Represents proportional gain, K i Indicates the integral gain, K d represents the differential gain, and Represents the error change rate of temperature and pressure at time t.
[0186] Numerical case analysis;
[0187] Material combination and process parameters: the ratio of polyurethane black material and white material: 1.4:1, temperature: 60°C; pressure: 100Pa, thickness: 150mm, state vector s = [1.4, 60, 100, 150];
[0188] PSO adjusts the material ratio and process parameters: the ratio of black polyurethane material to white polyurethane material is adjusted to 1.5:1; the temperature is adjusted to 65°C, and the action vector a=[1.5,65];
[0189] The current performance indicators are as follows: thermal insulation performance P1 = 0.85, bonding strength P2 = 0.9, durability P3 = 0.8, cost C = 0.7, production efficiency E = 0.9;
[0190] The target performance indicators are: thermal insulation performance P1 = 0.9, bonding strength P2 = 0.95, durability P3 = 0.85, cost C = 0.65, production efficiency E = 0.95;
[0191] Predicting performance indicators using DNN: Input: [1.4, 60, 100, 150] (material combination and process parameters); Output: [0.85, 0.9, 0.8, 0.7, 0.9] (performance indicator);
[0192] Use DDPG to dynamically adjust: According to the current state s = [1.4, 60, 100, 150], the action generated by the policy network is a t =[1.45,63], after adding exploration noise, we get a t=[1.48,64]; execute action a t =[1.48,64], observe the new status and rewards; new material ratio and process parameters s t+1 =[1.48,64,100,150], calculate reward: r=0.03;
[0193] PID control; target temperature r T (t) = 65°C, target pressure r P (t) = 105Pa; current actual temperature T(t) = 64°C, actual pressure P(t) = 102Pa; e T (t) = 65-64 = 1; e P (t) = 105-102 = 3;
[0194] After calculation, we get the temperature control output and pressure control output: K p =1.0, K i 0.1, K d =0.01
[0195] u T (t) = 1 + 0.1t; u P (t) = 3 + 0.3t;
[0196] Comparative Example 1
[0197] A method for preparing a composite building energy-saving thermal insulation board, specifically comprising the following preparation steps:
[0198] Step 1: Coat the 2.0 mm thick inner layer cement fiberglass felt 1 with epoxy resin adhesive, and adhere the 50 mm thick rock wool board 3 on the epoxy resin adhesive;
[0199] Step 2: Then put it into the laminator, the lamination temperature is 55°C, the conveying speed is 4.5m / min, and it is at the entrance of the laminator;
[0200] Step 3: Under normal temperature and pressure conditions, coat the outer surface of the material obtained in step 2 with an outer layer of cement fiberglass felt, then place it in a fume hood for volatilization at room temperature for 24 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
[0201] Performance Testing:
[0202] Waterproof performance: According to GB / T30805-2014, first use a balance to weigh the sample mass as M1, immerse the sample in a water tank, take it out after immersion for 24 hours and drain it for 10 minutes, weigh the sample mass as M2, and the sample size is 2mm×2mm×2mm; the water absorption calculation formula is:
[0203]
[0204] Among them, W b is the water absorption (kg / m 2 ); M1 is the mass of the sample before immersion (kg); M2 is the mass of the sample after immersion for 24 hours (kg); A is the lower surface area of the sample (m 2 ).
[0205] Compressive strength: According to GB / T 5486-2008, the sample is placed on a universal testing machine and the load is increased at a speed of (10+1) mm / min to cause structural damage to the sample or produce a 5% compression deformation. The failure load or the load when the sample is compressed and deformed by 5% is recorded; the results are shown in Table 1:
[0206] Table 1
[0207] Compressive strength / KPa Example 1 762 Example 2 766 Example 3 758 Comparative Example 1 621
[0208] Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.
[0209] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0210] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A composite building energy-saving thermal insulation board, characterized in that: It comprises a cement glass fiber felt (1), a thermal insulation layer (2), a rock wool board (3), a polyurethane board (4) and a polyurethane adhesive (5); A polyurethane adhesive (5) is coated on the surface of the cement glass fiber felt (1), a rock wool board (3) is adhered to the polyurethane adhesive (5), an upper polyurethane board (4) is arranged on the rock wool board (3) by using a high pressure machine, and a thermal insulation layer (2) is covered on the surface of the polyurethane board (4).
2. A composite building energy-saving thermal insulation board according to claim 1, characterized in that: Black polyurethane material and white polyurethane material are poured on the rock wool board (3) and foamed to form a polyurethane board (4), wherein the mass ratio of the black polyurethane material to the white polyurethane material is (1.2-1.6):
1.
3. The composite building energy-saving thermal insulation board according to claim 1, characterized in that: The preparation method of the thermal insulation layer (2) is as follows: adding SiO2 aerogel to anhydrous ethanol, stirring, and then adding KH550 and Tween 80 until they are evenly mixed, thereby obtaining the thermal insulation layer (2).
4. A method for preparing a composite building energy-saving thermal insulation board according to any one of claims 1 to 3, characterized in that: The specific preparation steps include: Step 1: Coat a 2.0 mm thick cement glass fiber felt (1) with a polyurethane adhesive (5), and adhere a 50 mm thick rock wool board (3) on the polyurethane adhesive (5); Step 2: Then put it into a laminator, and at the entrance of the laminator, use a high pressure machine to pour polyurethane black material and white material on the rock wool board (3), and foam to form a 150 mm thick polyurethane board (4); Step 3: Under normal temperature and pressure conditions, place the material obtained in step 2 in an insulation layer (2) for infiltration treatment, then place it in a fume hood for volatilization at room temperature for 24 to 30 hours, and finally place it in a drying oven at normal pressure and 50°C for drying for 24 hours to obtain a composite insulation board.
5. The method for preparing a composite building energy-saving thermal insulation board according to claim 4, characterized in that: The preparation method of the polyurethane adhesive (5) is: S1. Place 15-20 parts of polyether polyol in a drying oven at 120-140°C for 2-3 hours, then take it out and cool it down to 70-80°C; S2, mixing the dried polyether polyol with 10-15 parts of isophorone diisocyanate, placing in a dryer, heating to 80-90° C., and reacting for 1.5-2.5 hours; S3, adding 0.03-0.05 parts of catalyst, reacting for 0.5-1.5 hours, and then adding 0.1-0.2 parts of triethylene glycol; S4, mixing the product in S3 with 3-5 parts of castor oil, and placing the mixture in a drying oven at 120-130° C. for dehydration and drying; S5, pour in 0.4-1 part of chain extender, stir for 8-10 minutes until it is evenly mixed, add 0.05-0.1 part of auxiliary agent, continue stirring, then place in a degassed environment and let stand for 5-10 minutes, and introduce nitrogen again; S6. Place the product in S5 in a drying oven at 120° C. and cure for 3 to 4 hours to obtain a polyurethane adhesive.
6. The method for preparing a composite building energy-saving thermal insulation board according to claim 5, characterized in that: The catalyst in S3 is a mixture of stannous octoate and dibutyltin dilaurate.
7. The method for preparing a composite building energy-saving thermal insulation board according to claim 5, characterized in that: The auxiliary agents in S5 are dimethyl carbonate and diphenylmethane 4,4-diisocyanate, and the two are added sequentially.
8. The method for preparing a composite building energy-saving thermal insulation board according to claim 4, characterized in that: The lamination temperature in step 2 is 55-65° C., and the conveying speed during the lamination process is 4.5-5.5 m / min.
9. The method for preparing a composite building energy-saving thermal insulation board according to claim 4, characterized in that: Steps 1 to 3 also include using intelligent control algorithms to optimize the production process, which specifically include the following steps: Step 1: Define optimization objectives and variables: Comprehensive data collection through sensors: Temperature sensor: monitors temperature changes during the production process; Pressure sensor: measures the pressure inside the production equipment; Humidity sensor: monitors ambient humidity; Thickness sensor: measures the thickness of each layer of the insulation board; Flow sensor: monitors the flow ratio of black and white polyurethane materials Thermal insulation performance test: Use heat conduction equipment to measure the thermal insulation effect; Bond strength test: Measure the bond strength using a tensile testing machine; Durability testing: Conduct accelerated aging tests to measure the durability of the product; Raw material consumption: record the consumption of polyurethane black material, white material and other raw materials; Optimization goals: maximize thermal insulation performance P1, maximize bonding strength P2, maximize durability P3, minimize cost C, and maximize production efficiency E; Variable: material ratio x1, x2, ..., x n ; Process parameters y1, y2, ..., y m ; Set the particle swarm size N, the initial position and velocity of each particle; x i =(x i1 ,x i2 ,...,x in ),v i =(v i1 ,v i2 ,...,v in ),i=1,2,...,N Among them, x in is the position of the ith particle in the nth dimension; v in is the current velocity of the ith particle in the nth dimension; calculate the fitness function value f(x i ); f(x i )=w1P1+w2P2+w3P3-w4C+w5E Among them, w1, w2, w3, w4, w5 are the weights of each target; Update individual optimal and global optimal positions; p i =min(f(x i )),g=min(f(p i )) Among them, p i The individual optimal position of the i-th particle, g the global optimal position, that is, the optimal position among all particles, f(p i ) i The fitness function value of each particle, min(·) takes the minimum value function; Update speed and location: v ij (t+1)=wv ij (t)+c1r1[p ij -x ij (t)]+c2r2[g j -x ij (t)] x ij (t+1)=x ij (t)+v ij (t+1) Among them, w is the inertia weight, c1, c2 are learning factors, r1, r2 are random numbers, v ij (t+1) and x ij (t+1) is the velocity and position of the i-th particle in the j-th dimension at the next time step, v ij (t) is the velocity of the i-th particle in the j-th dimension at the current step, p ij is the individual optimal velocity of the ith particle in the jth dimension, x ij (t) is the current position of the ith particle in the jth dimension, g j is the global optimal position of all particles in the jth dimension, t represents the current time step, and t+1 represents the next step after the current step; Step 2: Use deep neural network DNN to predict the performance indicators of different combinations and use it as the objective function of PSO: Input layer: 2 neurons, corresponding to material ratio and process parameters; Hidden layer: 128 hidden layers are set, each hidden layer contains 8 neurons, and the activation function is ReLU; Output layer: 2 neurons, corresponding to performance indicators; Step 3: Use deep deterministic policy gradient (DDPG) to dynamically adjust parameters in the production process, and use PID controller for fine-tuning to ensure the stability of the production process. State s: material combination and process parameters; Action a: Adjust material ratio and process parameters; Reward r: Define the reward function based on product performance and cost; Select an action: in, is the exploration noise at time step t, a t is the action at time step t, s t is the state at time step t, θ μ represents the weight of the target network, μ(s t |θ μ ) indicates that the policy network is based on the current state s t and its weight θ μ Generate actions; perform actions and observe rewards and new states; r t =R(s t ,a t ),s t+1 =f(s t ,a t ) Among them, r t Indicates that action a is executed at time step t t Reward obtained after t , a t ) represents the reward function, according to the current state s t and action a i Calculate Rewards;s t+1 Indicates that action a is executed at time step t t The new state after f(s t , a t ) represents the state transition function, which describes the transition from state s t Transfer to new state s t+1 process; Storage experience: D←D∪(s t ,a t ,r t ,s t+1 ) (s i ,a i ,r i ,s i+1 )~D Where D is the experience replay pool, which stores state transition samples; (s t , a t , r t ,s t+1 ) is the empirical sample observed at time step t; (s i , a i , r i ,s i+1 ) represents an experience sample sampled from the experience replay pool; Update the Critic network: y i =r i +γQ′(s i+1 ,μ′(s i+1 |θ μ′ )|θ Q′ ) Among them, y i Represents the target value, used to update the Critic network, r i represents the reward; γ represents the discount factor, which measures the importance of future rewards; Q′(s i+1 ,μ′(s i+1 |θ μ′ )|θ Q′ ) represents the output of the target Critic network, and evaluates the new state s i+1 The action μ′(s) generated by the execution policy network is i+1 |θ μ′ ) value; Q′ represents the target Critic network, which is a delayed update version of the Critic network used for stable training; μ′(s i+1 |θ μ′ ) The target Actor network generates s according to the state i+1 The target Actor network is a delayed updated version of the Actor network, θ μ′ Represents the parameters of the target Actor network; θ Q′ Represents the parameters of the target Critic network; Among them, L represents the loss function of the Critic network, Q(s i , a i |θ Q ) represents the output of the Critic network, evaluating the state s i Next, perform action a i The value of N represents the batch size, θ Q It is the Critic network; in, Represents the gradient of the policy network, which is used to update the Actor network parameters θ μ ; Represents the gradient of the Critic network relative to action a, indicating that in state s i Next, take action a = μ(s i ) is the gradient of the value function when Q and θ μ are the parameters of the Critic network and the Actor network respectively; μ(s i ) indicates that the policy network is based on the state s i The actions generated; The network is related to the parameter θ μ The gradient of Soft update target network: θ′←τθ+(1-τ)θ′ Among them, θ represents the parameters of the current network, θ′ represents the parameters of the target network, and τ represents the soft update coefficient; Step 4: Adjust the parameters of the production equipment through the PID controller; e T (t)=r T (t)-T(t) e P (t)=r P (t)-P(t) Among them, e T (t) and e P (t) represents the temperature error and pressure error at time t, r T (t) and r P (t) target temperature and target pressure at time t, T(t) and P(t) actual temperature and actual pressure at time t, u T (t) and u P (t) Control output at time t, used to adjust temperature and pressure, K p Represents proportional gain, K i Indicates the integral gain, K d represents the differential gain, and Represents the error change rate of temperature and pressure at time t.
Citation Information
Cited By
Multi-parameter regulation and control method and system for thermal insulation plate production and medium
CN121008553A
Multi-parameter regulation method and system for thermal insulation board production and medium
CN121008553B
