Krill ship waste heat temperature control method, model and equipment and storage medium

Abstract

The invention discloses a krill ship waste heat temperature control method, model and device and a storage medium, and the method comprises the steps: collecting the waste heat power and waste heat pressure of a krill ship waste gas pipeline, and generating a time-related waste heat sequence according to the waste heat power and waste heat pressure; the method comprises the steps of decomposing a time-dependent waste heat sequence into a preset number of waste heat subsequences through a VMD algorithm model, optimizing model parameters of an Elman neural network through a GWO algorithm model according to the preset number of waste heat subsequences, and performing temperature prediction through the optimized Elman neural network. The preset temperature control model achieves target krill ship waste heat temperature control through the PID controller based on the predicted waste heat temperature and the actual waste heat temperature. According to the method, better temperature control and shorter training time are realized through the VMD-GWO-Elman-PID combined model, so that the waste heat temperature control precision of the krill ship is remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of temperature control, and in particular relates to a krill ship waste heat temperature control method, model, device and storage medium. Background Art

[0002] At present, magnetic levitation chillers are widely used in krill ships. The commonly used control method is the traditional PID control algorithm. The difference between the actual outlet water temperature of the krill ship exhaust pipe and the set temperature is taken as the control object, and the three parameters of proportional coefficient (P), integral coefficient (I) and differential coefficient (D) are used for adjustment. The traditional PID algorithm has a stable control effect on simple control systems, but for the magnetic levitation variable frequency chiller control system, because there are too many uncertain variables, it is difficult to describe the operation mode of each component of the unit through an accurate mathematical model, and the entire control feedback system has strong nonlinearity and high time lag, and there is a difficulty of non-minimum phase. The existence of non-minimum phase makes the system have a certain phase lag, which leads to a higher control feedback delay of the outlet water temperature of the krill ship exhaust pipe, which leads to poor control effect, especially the rapid response of the feedback change of the outlet water temperature of the krill ship exhaust pipe becomes poor, and the tracking calibration of the system will also be more difficult, which leads to a certain degree of calibration error when using the traditional PID algorithm, and thus the ideal control effect cannot be achieved.

[0003] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0004] The main purpose of the present invention is to provide a krill ship waste heat temperature control method, model, device and storage medium, aiming to solve the technical problem of how to improve the control accuracy of krill ship waste heat temperature.

[0005] To achieve the above object, the present invention provides a krill ship waste heat temperature control method, the krill ship waste heat temperature control method comprising:

[0006] collecting waste heat power and waste heat pressure from the exhaust gas pipeline of the krill vessel, and generating a time-related waste heat sequence according to the waste heat power and the waste heat pressure;

[0007] The time-dependent waste heat sequence is input into a preset temperature control model, wherein the preset temperature prediction model is constructed by combining a VMD algorithm model, a GWO algorithm model, an Elman neural network and a PID controller;

[0008] The time-related waste heat sequence is decomposed into a preset number of waste heat sub-sequences through the VMD algorithm model, and the model parameters of the Elman neural network are optimized through the GWO algorithm model according to the preset number of waste heat sub-sequences, and temperature prediction is performed through the optimized Elman neural network, so that the preset temperature control model achieves target waste heat temperature control through the PID controller based on the predicted waste heat temperature and the actual waste heat temperature.

[0009] Optionally, the step of generating a time-dependent waste heat sequence according to the waste heat power and the waste heat pressure comprises:

[0010] constructing a waste heat hotspot set according to the waste heat power and the waste heat pressure in accordance with time coding, and generating a waste heat distribution map according to the waste heat hotspot set;

[0011] Dividing a plurality of waste heat pressure intervals and a plurality of waste heat power intervals based on the waste heat distribution map;

[0012] The quartile points of waste heat pressure in each waste heat pressure interval and the quartile points of waste heat power in each waste heat power interval are determined respectively by using the quintile algorithm;

[0013] According to the quartiles of the waste heat pressure in each waste heat pressure interval, a pressure quantile interval corresponding to each waste heat pressure interval is constructed, and according to the quartiles of the waste heat power in each waste heat power interval, a power quantile interval corresponding to each waste heat power interval is constructed;

[0014] A time-dependent waste heat sequence is determined based on the pressure quantile intervals corresponding to each waste heat pressure interval and the power quantile intervals corresponding to each waste heat power interval.

[0015] Optionally, the step of determining the time-dependent waste heat sequence based on the pressure quantile intervals corresponding to each waste heat pressure interval and the power quantile intervals corresponding to each waste heat power interval comprises:

[0016] Determine abnormal pressure data outside the pressure quantile interval corresponding to each waste heat pressure interval, and calculate the waste heat pressure average value in the waste heat pressure interval corresponding to the abnormal pressure data, so that the waste heat pressure average value replaces the abnormal pressure data;

[0017] Determine abnormal power data outside the power quantile interval corresponding to each waste heat power interval, and calculate the waste heat power average value in the waste heat power interval corresponding to the abnormal power data, so that the waste heat power average value replaces the abnormal power data;

[0018] Outputting the waste heat pressure data and waste heat power data corresponding to the waste heat distribution diagram in a time-coded order;

[0019] generating a waste heat pressure and power sequence according to the waste heat pressure data and the waste heat power data;

[0020] The waste heat pressure power sequence is normalized to obtain a time-dependent waste heat sequence.

[0021] Optionally, the step of decomposing the time-dependent waste heat sequence into a preset number of waste heat sub-sequences by using a VMD algorithm model comprises:

[0022] Constructing a modal broadband constrained variational problem through a VMD algorithm model according to the time-dependent waste heat sequence;

[0023] The modal broadband constrained variational problem is converted into a modal broadband unconstrained variational problem by using a quadratic penalty factor and a Lagrange multiplication operator;

[0024] The modal broadband unconstrained variational problem is solved by an alternating direction method of multipliers and Fourier transform to obtain a preset number of waste heat subsequences.

[0025] Optionally, the step of optimizing the model parameters of the Elman neural network through the GWO algorithm model according to the preset number of waste heat subsequences includes:

[0026] The GWO algorithm model is optimized based on the convergence factor of sinusoidal changes to obtain the GWO correction expression.

[0027] Constructing a temperature search formula according to the GWO correction expression;

[0028] Obtaining an optimal temperature measurement value by using the temperature search formula according to the preset number of waste heat subsequences;

[0029] constructing a waste heat temperature data set according to the temperature measurement values ​​at a preset time, the preset number of waste heat subsequences and the optimal temperature measurement value;

[0030] The model parameters of the Elman neural network are optimized based on the waste heat temperature data set.

[0031] Optionally, the step of achieving target waste heat temperature control through a PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature includes:

[0032] determining a temperature difference signal according to a target waste heat temperature and the actual waste heat temperature;

[0033] The target waste heat temperature is controlled by a PID control algorithm according to the temperature difference signal and the predicted waste heat temperature.

[0034] In addition, to achieve the above purpose, the present invention also proposes a preset temperature control model based on the krill ship waste heat temperature control method, and the preset temperature prediction model is constructed by combining the VMD algorithm model, the GWO algorithm model, the Elman neural network and the PID controller:

[0035] The VMD algorithm model is used to decompose the time-dependent waste heat sequence into a preset number of waste heat subsequences;

[0036] The GWO algorithm model is used to optimize the model parameters of the Elman neural network according to the preset number of waste heat subsequences, so that the optimized Elman neural network performs temperature prediction;

[0037] The PID controller is used to achieve target waste heat temperature control through a PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature.

[0038] In addition, to achieve the above-mentioned purpose, the present invention also proposes a krill vessel waste heat temperature control device, the device comprising: a memory, a processor, and a krill vessel waste heat temperature control program stored in the memory and executable on the processor, the krill vessel waste heat temperature control program being configured to implement the steps of the krill vessel waste heat temperature control method as described above.

[0039] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a krill ship waste heat temperature control program is stored. When the krill ship waste heat temperature control program is executed by a processor, the steps of the krill ship waste heat temperature control method as described above are implemented.

[0040] The present invention first collects the waste heat power and waste heat pressure of the krill ship exhaust gas pipeline, and generates a time-related waste heat sequence according to the waste heat power and waste heat pressure, and then decomposes the time-related waste heat sequence into a preset number of waste heat subsequences through the VMD algorithm model, and then optimizes the model parameters of the Elman neural network through the GWO algorithm model according to the preset number of waste heat subsequences, and performs temperature prediction through the optimized Elman neural network to output the predicted waste heat temperature, and finally realizes the target waste heat temperature control through the PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature. The present invention obtains a series of finite stationary subsequence components based on the time-related waste heat sequence through the VMD algorithm model, and clusters the subsequence components using the GWO algorithm model, superimposes each type to reduce the training time, and then trains the neural network for prediction with the clustered sequence, superimposes and denormalizes the prediction results of each subsequence to obtain the final prediction result, and outputs it to the PID for temperature control until the actual temperature is consistent with the target temperature, thereby achieving an ideal control effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a structural schematic diagram of a krill ship waste heat temperature control device in a hardware operating environment involved in an embodiment of the present invention;

[0042] Figure 2 It is a flow chart of the first embodiment of the waste heat temperature control method of a krill ship of the present invention;

[0043] Figure 3 It is a structural block diagram of the first embodiment of the preset temperature control model of the present invention.

[0044] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0045] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0046] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a krill ship waste heat temperature control device in the hardware operating environment involved in an embodiment of the present invention.

[0047] like Figure 1 As shown, the krill ship waste heat temperature control device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM), or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory 1005 may also be a storage system independent of the aforementioned processor 1001.

[0048] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the waste heat temperature control device for krill vessels, and may include more or less components than those shown in the figure, or combine certain components, or arrange the components differently.

[0049] like Figure 1As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a krill boat waste heat temperature control program.

[0050] exist Figure 1 In the krill ship waste heat temperature control device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the krill ship waste heat temperature control device of the present invention can be set in the krill ship waste heat temperature control device, and the krill ship waste heat temperature control device calls the krill ship waste heat temperature control program stored in the memory 1005 through the processor 1001, and executes the krill ship waste heat temperature control method provided in the embodiment of the present invention.

[0051] The embodiment of the present invention provides a method for controlling the temperature of waste heat from a krill ship, referring to Figure 2 , Figure 2 It is a schematic flow chart of the first embodiment of the waste heat temperature control method of a krill vessel according to the present invention.

[0052] In this embodiment, the krill ship waste heat temperature control method includes the following steps:

[0053] Step S10: collecting the waste heat power and waste heat pressure of the krill ship exhaust pipe, and generating a time-related waste heat sequence according to the waste heat power and the waste heat pressure.

[0054] It is easy to understand that the execution subject of this embodiment can be a processor with functions such as data processing, network communication and program running, or other computer devices with similar functions, etc., and this embodiment is not limited.

[0055] It should be noted that the target waste heat temperature to be controlled is input, the water temperature is detected by the sensor, and the waste heat power p1 and waste heat pressure p2 of the krill ship exhaust pipe are collected.

[0056] Furthermore, a processing method for generating a time-dependent waste heat sequence according to waste heat power and waste heat pressure is as follows: constructing a waste heat hotspot set according to the waste heat power and waste heat pressure according to time coding, and generating a waste heat distribution map according to the waste heat hotspot set; dividing a plurality of waste heat pressure intervals and a plurality of waste heat power intervals based on the waste heat distribution map; determining the quartile points of the waste heat pressure in each waste heat pressure interval and the quartile points of the waste heat power in each waste heat power interval respectively through a quintile algorithm; constructing a pressure quantile interval corresponding to each waste heat pressure interval according to the quartile points of the waste heat pressure in each waste heat pressure interval, and constructing a power quantile interval corresponding to each waste heat power interval according to the quartile points of the waste heat power in each waste heat power interval; determining a time-dependent waste heat sequence based on the pressure quantile interval corresponding to each waste heat pressure interval and the power quantile interval corresponding to each waste heat power interval.

[0057] In the specific implementation, the waste heat power p1 and waste heat pressure p2 are encoded according to the sampling time t to form a point set (p1, p2) (i.e., a waste heat point set), and the distribution diagram of the original data (p1, p2) (i.e., a waste heat distribution diagram) is drawn. Based on the waste heat distribution diagram, the waste heat pressure p2 is divided into intervals Δp2, and the four quantiles M1, M2, M3, and M4 of the waste heat pressure p2 are found in each Δp2 interval using the quintile algorithm in turn, and N=M4-M1 is calculated to construct the pressure quantile interval [M1-1.5N, M4+1.5N]; Based on the waste heat distribution diagram, the waste heat power p1 is divided into intervals Δp1, and the four quantiles M1, M2, M3, and M4 of the waste heat power p1 are found in each Δp1 interval using the quintile algorithm in turn, and N=M4-M1 is calculated to construct the power quantile interval [M1-1.5N, M4+1.5N].

[0058] Furthermore, a processing method for determining a time-correlated waste heat sequence based on the pressure quantile interval corresponding to each waste heat pressure interval and the power quantile interval corresponding to each waste heat power interval is as follows: determining abnormal pressure data outside the pressure quantile interval corresponding to each waste heat pressure interval, and calculating the waste heat pressure average value in the waste heat pressure interval corresponding to the abnormal pressure data, so that the waste heat pressure average value replaces the abnormal pressure data; determining abnormal power data outside the power quantile interval corresponding to each waste heat power interval, and calculating the waste heat power average value in the waste heat power interval corresponding to the abnormal power data, so that the waste heat power average value replaces the abnormal power data; outputting the waste heat pressure data and waste heat power data corresponding to the waste heat distribution diagram in a time-coded order; generating a waste heat pressure power sequence according to the waste heat pressure data and the waste heat power data; and normalizing the waste heat pressure power sequence to obtain a time-correlated waste heat sequence.

[0059] In the specific implementation, the abnormal pressure data outside the pressure quantile interval corresponding to the interval Δp2 is deleted, and the average value of the waste heat pressure in each interval Δp2 is taken to replace the abnormal pressure data outside the pressure quantile interval; the abnormal power data outside the power quantile interval corresponding to the interval Δp1 is deleted, and the average value of the waste heat power in each interval Δp1 is taken to replace the abnormal pressure data outside the power quantile interval.

[0060] Furthermore, the correct waste heat pressure data and waste heat power data are output in the order of time coding, and a waste heat pressure power sequence pa is generated according to the waste heat pressure data and the waste heat power data. The sequence pa is normalized to obtain the sequence Pi (i.e., the time-related waste heat sequence).

[0061] Step S20: inputting the time-dependent waste heat sequence into a preset temperature control model, wherein the preset temperature prediction model is constructed by combining a VMD algorithm model, a GWO algorithm model, an Elman neural network and a PID controller;

[0062] Step S30: Decomposing the time-related waste heat sequence into a preset number of waste heat sub-sequences through the VMD algorithm model, optimizing the model parameters of the Elman neural network through the GWO algorithm model according to the preset number of waste heat sub-sequences, and performing temperature prediction through the optimized Elman neural network, so that the preset temperature control model achieves target waste heat temperature control through the PID controller based on the predicted waste heat temperature and the actual waste heat temperature.

[0063] In the specific implementation, compared with the traditional control algorithm, the combination model of variational mode decomposition (VMD)-Grey Wolf Optimizer (GWO)-Elman neural network-PID controller (i.e., the preset temperature control model) is characterized by stronger self-learning ability. By continuously adjusting the changes in various parameters during the operation of the control model, it controls and adjusts the unit operation with stronger nonlinear mapping ability and adaptive control adjustment ability.

[0064] Compared with the control effect of traditional PID algorithm, the use of VMD-GWO-ELMAN-PID combined model to control the outlet water temperature of the krill ship exhaust pipe can further shorten the time for the outlet water temperature of the krill ship exhaust pipe to reach the target temperature.

[0065] After adopting the VMD-GWO-ELMAN-PID combined model, the overshoot of the outlet water temperature of the krill ship exhaust pipe is reduced, and the temperature fluctuation is reduced. When the traditional PID algorithm is used, the outlet water temperature will fluctuate immediately when interference occurs. After a period of adjustment, it will recover, but the deviation still exists. The VMD-GWO-ELMAN-PID combined model can stabilize the outlet water temperature at the set value.

[0066] In this embodiment, the time-dependent waste heat sequence is decomposed into a preset number of waste heat sub-sequences through a VMD algorithm model.

[0067] Furthermore, a processing method for decomposing a time-dependent waste heat sequence into a preset number of waste heat sub-sequences through the VMD algorithm model is to construct a modal broadband constrained variational problem according to the time-dependent waste heat sequence through the VMD algorithm model; convert the modal broadband constrained variational problem into a modal broadband unconstrained variational problem through a quadratic penalty factor and a Lagrange multiplication operator; and solve the modal broadband unconstrained variational problem through the alternating direction method of multipliers and Fourier transform to obtain a preset number of waste heat sub-sequences.

[0068] In the specific implementation, the modal broadband constrained variational problem is constructed through the VMD algorithm model according to the time-related waste heat sequence to ensure that the decomposed waste heat subsequence is a modal component with a finite bandwidth and a center frequency. At the same time, the sum of the estimated bandwidths of each mode is minimized, and the constraint conditions are established.

[0069] In this embodiment, the time-dependent waste heat sequence is written as a time series y(t), that is, y(t) is the sum of each modal component. First, the analytical spectrum of each mode mk(t) is obtained by Hilbert transform, where the one-sided spectrum Amk(t) is:

[0070]

[0071] Here, δ represents the Dirac distribution and j represents the imaginary unit.

[0072] By mixing the center frequency into the decomposed time series, the spectrum will be adjusted to the baseband, i.e., Amk(t).

[0073] Among them, ω k represents the kth center frequency; the bandwidth of each modal signal is estimated by using the Gaussian smoothing method of the demodulated signal, which can solve the constrained variational problem;

[0074] The objective function is expressed as:

[0075]

[0076] Among them, {m k}={m 1 m 2 ...m k},{ω k}={ω 1 ω 2 ... k}, K represents the number of modes to be decomposed (i.e., a preset number of waste heat subsequences need to be decomposed), m k represents the kth modal component, ω k represents the kth center frequency, δ(t) represents the Dirac function, and * represents the convolution operation.

[0077] The quadratic penalty factor α and the Lagrange multiplication operator λ(t) are introduced to transform the constrained variational problem into an unconstrained variational problem:

[0078] α ensures the reconstruction accuracy of the signal, and λ(t) maintains strict constraints:

[0079] Using the alternating direction method of multipliers and Fourier transform to solve the above variational problem, the best solution to the augmented Lagrangian function expression is given by the alternating update optimization The expression after alternating optimization iteration is as follows:

[0080]

[0081] in, Respectively is the Fourier isometric transform of , and τ represents the noise tolerance.

[0082] It should also be noted that initialization is required before starting to run And set the maximum number of iterations N, then update right Perform Fourier transform and we can get: m 1 (t),m 2 (t),m 3 (t),..., (i.e., a preset number of waste heat subsequences u(t)).

[0083] The model parameters of the Elman neural network are optimized through the GWO algorithm model according to a preset number of waste heat subsequences, and the temperature is predicted by the optimized Elman neural network to output the predicted waste heat temperature.

[0084] Furthermore, the GWO algorithm model is optimized based on the convergence factor of the sinusoidal law to obtain the GWO correction expression; a temperature search formula is constructed according to the GWO correction expression; and the optimal temperature measurement value (i.e., X) is obtained through the temperature search formula according to a preset number of waste heat subsequences. α , X β and X χ ); construct a waste heat temperature data set according to the temperature measurement value X(t) at a preset time, a preset number of waste heat subsequences and the optimal temperature measurement value; and optimize the model parameters of the Elman neural network based on the waste heat temperature data set.

[0085] The GWO correction expression is:

[0086]

[0087] Initialize wolf population parameters. Set parameters such as gray wolf population size, search space, and maximum number of iterations;

[0088] Calculate the fitness value of each individual gray wolf in the population;

[0089] According to the fitness value, the first three L subsequences among the preset number of waste heat subsequences are recorded as the three gray wolves with the best performance and are recorded as α, β and χ.

[0090] Temperature search formula:

[0091] D=rand{0,2}X P (txt)

[0092] X(t+1)=X P (t)-rand(-2,2)*rand(0,2)

[0093]

[0094] D i =rand{0,2}X i -X

[0095] In the formula, X P (t) is the optimal value of the current solution, and X(t) is the temperature measurement value of the current solution, i.e., time t. α , X β and X χ These are the values ​​of the first three wolves obtained through GWO optimization.

[0096] In this embodiment, it is also necessary to establish an Elman neural network:

[0097] x(k)=f(w 1 x c (k)+w 2 (x(k-1))+b s ), y k =g(w 3 x(k)+b t )

[0098] x c (k) = [X(t),u(t),X α ,X β ,X χ ]

[0099] Among them, x c (k) is the waste heat temperature dataset, x(k) is the state vector of the hidden layer, f(·) is the activation function ReLU, and w 1 is the weight matrix from input to hidden layer, w 2 is the weight matrix of the feedback connection from the hidden layer state at the previous moment to the hidden layer state at the current moment. s is the bias vector of the hidden layer. kis the output (i.e., predicted waste heat temperature), g(·) is the activation function of the output layer, and w 3 is the weight matrix from the hidden layer to the output layer, b t is the bias vector of the output layer.

[0100] It should be noted that the model parameters of the Elman neural network are optimized based on the waste heat temperature data set so that the optimized Elman neural network outputs the predicted waste heat temperature.

[0101] The target waste heat temperature is controlled by a PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature.

[0102] Further, a temperature difference signal is determined according to the target waste heat temperature and the actual waste heat temperature; and the target waste heat temperature is controlled by a PID control algorithm according to the temperature difference signal and the predicted waste heat temperature.

[0103] In the specific implementation, the PID controller output x uses the following incremental PID algorithm:

[0104] y k =y(k-1)+k p [e(k)-e(k-1)]+k i e(k)+k d [e(k)-2e(k-1)+e(k-2)]

[0105] e(k)=T r (k)-T o (k)

[0106] In the formula, k p ,k i ,k d are the proportional, integral and differential control parameters of VMD-GWO-ELMAN-PID controller, k is time, T r (k) is the target temperature, T o (k) is the actual temperature, and e(k) is the temperature difference signal.

[0107] It should also be noted that the PID controller transmits the signal to the inverter to adjust the compressor speed, and finally the waste heat pipe transmits the water temperature. The water temperature parameter is output, and if it is equal to the target temperature, the current working environment is maintained. If it is not equal, the temperature is transmitted back to the PID controller. The above cycle is repeated until the temperature is equal.

[0108] In this implementation, the waste heat power and waste heat pressure of the krill ship exhaust gas pipeline are first collected, and a time-related waste heat sequence is generated according to the waste heat power and waste heat pressure, and then the time-related waste heat sequence is decomposed into a preset number of waste heat subsequences through the VMD algorithm model, and then the model parameters of the Elman neural network are optimized through the GWO algorithm model according to the preset number of waste heat subsequences, and the temperature is predicted through the optimized Elman neural network to output the predicted waste heat temperature, and finally the target waste heat temperature is controlled through the PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature. In this embodiment, a series of finite stationary subsequence components are obtained through the VMD algorithm model based on the time-related waste heat sequence, and the subsequence components are clustered using the GWO algorithm model, and the various types are superimposed to reduce the training time, and the clustered sequences are used to train the neural network for prediction, and the prediction results of each subsequence are superimposed and denormalized to obtain the final prediction result, which is output to the PID for temperature control until the actual temperature is consistent with the target temperature, thereby improving the temperature control accuracy.

[0109] Reference Figure 3 , Figure 3 It is a structural block diagram of the first embodiment of the preset temperature control model of the present invention.

[0110] like Figure 3 As shown, the preset temperature control model based on the krill ship waste heat temperature control method proposed in the embodiment of the present invention includes a VMD algorithm model, a GWO algorithm model, an Elman neural network and a PID controller:

[0111] The VMD algorithm model 3001 is used to decompose the time-related waste heat sequence into a preset number of waste heat sub-sequences;

[0112] The GWO algorithm model 3002 is used to optimize the model parameters of the Elman neural network 3003 according to the preset number of waste heat subsequences, so that the optimized Elman neural network performs temperature prediction;

[0113] The PID controller 3004 is used to achieve target waste heat temperature control through a PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature.

[0114] Other embodiments or specific implementations of the preset temperature control model of the present invention can refer to the above-mentioned method embodiments, which will not be described in detail here.

[0115] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system 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 system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0116] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0117] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0118] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A krill ship waste heat temperature control method, characterized in that: The krill ship waste heat temperature control method comprises the following steps: collecting waste heat power and waste heat pressure from the exhaust gas pipeline of the krill vessel, and generating a time-related waste heat sequence according to the waste heat power and the waste heat pressure; The time-dependent waste heat sequence is input into a preset temperature control model, wherein the preset temperature prediction model is constructed by combining a VMD algorithm model, a GWO algorithm model, an Elman neural network and a PID controller; The time-related waste heat sequence is decomposed into a preset number of waste heat sub-sequences through the VMD algorithm model, and the model parameters of the Elman neural network are optimized through the GWO algorithm model according to the preset number of waste heat sub-sequences, and temperature prediction is performed through the optimized Elman neural network, so that the preset temperature control model achieves target waste heat temperature control through the PID controller based on the predicted waste heat temperature and the actual waste heat temperature.

2. The method according to claim 1, characterized in that The step of generating a time-dependent waste heat sequence according to the waste heat power and the waste heat pressure comprises: constructing a waste heat hotspot set according to the waste heat power and the waste heat pressure in accordance with time coding, and generating a waste heat distribution map according to the waste heat hotspot set; Dividing a plurality of waste heat pressure intervals and a plurality of waste heat power intervals based on the waste heat distribution map; The quartile points of waste heat pressure in each waste heat pressure interval and the quartile points of waste heat power in each waste heat power interval are determined respectively by using the quintile algorithm; According to the quartiles of the waste heat pressure in each waste heat pressure interval, a pressure quantile interval corresponding to each waste heat pressure interval is constructed, and according to the quartiles of the waste heat power in each waste heat power interval, a power quantile interval corresponding to each waste heat power interval is constructed; A time-dependent waste heat sequence is determined based on the pressure quantile intervals corresponding to each waste heat pressure interval and the power quantile intervals corresponding to each waste heat power interval.

3. The method according to claim 2, characterized in that The step of determining the time-dependent waste heat sequence based on the pressure quantile intervals corresponding to each waste heat pressure interval and the power quantile intervals corresponding to each waste heat power interval comprises: Determine abnormal pressure data outside the pressure quantile interval corresponding to each waste heat pressure interval, and calculate the waste heat pressure average value in the waste heat pressure interval corresponding to the abnormal pressure data, so that the waste heat pressure average value replaces the abnormal pressure data; Determine abnormal power data outside the power quantile interval corresponding to each waste heat power interval, and calculate the waste heat power average value in the waste heat power interval corresponding to the abnormal power data, so that the waste heat power average value replaces the abnormal power data; Outputting the waste heat pressure data and waste heat power data corresponding to the waste heat distribution diagram in a time-coded order; generating a waste heat pressure and power sequence according to the waste heat pressure data and the waste heat power data; The waste heat pressure power sequence is normalized to obtain a time-dependent waste heat sequence.

4. The method according to any one of claims 1 to 3, characterized in that: The step of decomposing the time-dependent waste heat sequence into a preset number of waste heat sub-sequences by using the VMD algorithm model comprises: Constructing a modal broadband constrained variational problem through the VMD algorithm model according to the time-dependent waste heat sequence; The modal broadband constrained variational problem is converted into a modal broadband unconstrained variational problem by using a quadratic penalty factor and a Lagrange multiplication operator; The modal broadband unconstrained variational problem is solved by an alternating direction method of multipliers and Fourier transform to obtain a preset number of waste heat subsequences.

5. The method according to claim 4, characterized in that The step of optimizing the model parameters of the Elman neural network through the GWO algorithm model according to the preset number of waste heat subsequences includes: Optimizing the GWO algorithm model based on the convergence factor of sinusoidal changes to obtain a GWO correction expression; Constructing a temperature search formula according to the GWO correction expression; Obtaining an optimal temperature measurement value by using the temperature search formula according to the preset number of waste heat subsequences; constructing a waste heat temperature data set according to the temperature measurement values ​​at a preset time, the preset number of waste heat subsequences and the optimal temperature measurement value; The model parameters of the Elman neural network are optimized based on the waste heat temperature data set.

6. The method according to claim 5, characterized in that The step of achieving target waste heat temperature control by the PID controller based on the predicted waste heat temperature and the actual waste heat temperature comprises: determining a temperature difference signal according to a target waste heat temperature and the actual waste heat temperature; The target waste heat temperature is controlled by the PID controller according to the temperature difference signal and the predicted waste heat temperature.

7. A preset temperature control model based on the krill ship waste heat temperature control method, characterized in that: The preset temperature prediction model is constructed by combining VMD algorithm model, GWO algorithm model, Elman neural network and PID controller: The VMD algorithm model is used to decompose the time-dependent waste heat sequence into a preset number of waste heat subsequences; The GWO algorithm model is used to optimize the model parameters of the Elman neural network according to the preset number of waste heat subsequences, so that the optimized Elman neural network performs temperature prediction; The PID controller is used to achieve target waste heat temperature control through a PID control algorithm based on the predicted waste heat temperature and the actual waste heat temperature.

8. A waste heat temperature control device for a krill ship, characterized in that: The device comprises: a memory, a processor, and a krill vessel waste heat temperature control program stored in the memory and executable on the processor, wherein the krill vessel waste heat temperature control program is configured to implement the steps of the krill vessel waste heat temperature control method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores a krill vessel waste heat temperature control program, and when the krill vessel waste heat temperature control program is executed by the processor, the steps of the krill vessel waste heat temperature control method according to any one of claims 1 to 6 are implemented.

Images