Ship heat source coupling power generation system

Abstract

The invention discloses a ship heat source coupling power generation system, which utilizes scavenging air, cylinder sleeve water and flue gas waste heat to generate power through two cycles, and controls two circulation loops through a control device to improve the waste heat source utilization efficiency. The power generation system comprises a multi-waste-heat-source power generation device and a control device, the multi-waste-heat-source power generation device comprises a first circulation loop and a second circulation loop, and the evaporation pressure and the superheat degree in the first circulation loop and the second circulation loop are controlled through the control device; the control device predicts the power of a host and the temperature of a heat source by collecting wave and wind speed data and utilizing a convolutional neural network, and outputs the corresponding optimal evaporation pressure and the optimal superheat degree according to prediction. And the evaporation pressure and the superheat degree of the first circulation loop and the second circulation loop are controlled by controlling the rotating speed of the expansion machine and the rotating speed of the working medium pump in the first circulation loop and the second circulation loop through a nonlinear model prediction algorithm, so that the power generation device outputs the maximum power.

Description

Technical Field

[0001] This invention belongs to the field of ship power generation applications, and specifically relates to a ship heat source coupled power generation system. Background Technology

[0002] Most ocean-going vessels use internal combustion engines as their primary power source, which consume a large amount of fuel during operation. Only 30% to 50% of the energy in this fuel is effectively utilized; most of the heat is discharged as waste heat through scavenging air, cylinder liner water, and exhaust gases. The exhaust temperature of a ship's main engine is typically high, and a large amount of its heat energy is directly released into the atmosphere, resulting in energy waste and thermal pollution. Simultaneously, the cylinder liner water absorbs a significant amount of heat during engine cooling, but this heat is usually simply cooled before being discharged without being fully recovered and utilized.

[0003] Currently, some shipboard waste heat power generation devices exist. However, these devices can only recover heat from a single heat source, such as recovering only waste heat from flue gas. They cannot comprehensively utilize multiple heat sources, resulting in insufficient heat source utilization and low energy efficiency. Moreover, existing control methods are mostly simple, primarily based on real-time operating conditions, leading to lagging control strategies and reduced heat-to-electricity conversion efficiency. Furthermore, existing power generation device control systems do not consider factors such as sea wave conditions, wind speed, and real-time main engine power, failing to provide intelligent and precise control. This prevents the power generation device from operating at its optimal state, further impacting energy recovery and utilization efficiency. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, the present invention aims to provide a technical solution for a ship heat source coupled power generation system. This system generates electricity by simultaneously utilizing scavenging air, cylinder liner water, and waste heat from flue gas in two cycles. It also predicts main engine power and heat source temperature by collecting wave and wind speed data and combining them with a neural network intelligent control strategy to achieve multi-heat source coordinated and optimized power generation, thereby improving the utilization efficiency of waste heat sources.

[0005] To achieve the above objectives, the present invention can be implemented through the following specific technical solutions: The aforementioned ship heat source coupled power generation system includes a waste heat source power generation device and a control device. The waste heat source power generation device includes a first circulation loop and a second circulation loop. The evaporation pressure and superheat in the first and second circulation loops are controlled by the control device, enabling simultaneous power generation using three types of waste heat: scavenging air, cylinder liner water, and flue gas. The control device collects wave and wind speed data, uses a convolutional neural network to predict the main engine power and heat source temperature, outputs the corresponding optimal evaporation pressure and optimal superheat based on the prediction, and controls the expander speed and working fluid pump speed in the first and second circulation loops through a nonlinear model prediction algorithm to control the evaporation pressure and superheat in the first and second circulation loops, so that the power generation device outputs maximum power.

[0006] Furthermore, both the first and second circulation loops are based on organic Rankine cycle technology for power generation. The first circulation loop is used to recover waste heat from the ship's main engine flue gas and generate electricity using the flue gas. The second circulation loop is used to recover waste heat from the scavenging air and cylinder liner water and generate electricity using the cylinder liner water and scavenging air. The first and second circulation loops are connected to a plate heat exchanger, and the electricity output from the two circulation loops is fed into the ship's power grid through an inverter. Each evaporator outlet of the power generation system is equipped with a temperature sensor and a pressure sensor. An encoder is installed on the expander shaft to measure the expander speed and transmits the data to the control device via bus communication. The speed of the working fluid pump in the waste heat source power generation device is controlled by a frequency converter-driven motor, and the expander speed is controlled by the rectifier to control the current. The control signal is sent by the control device to the frequency converter and rectifier in the waste heat source power generation device via bus communication.

[0007] Furthermore, the first circulation loop includes a liquid storage tank I, a working fluid pump I, a shell-and-tube heat exchanger, a screw expander I, and a plate heat exchanger connected in series. The screw expander I is also connected in series with a generator I, a rectifier I, and an inverter I. The working fluid pump I is connected to a motor I, and the motor I is driven by a frequency converter I to draw subcooled liquid from the liquid storage tank I and pump it into the shell-and-tube heat exchanger. In the shell-and-tube heat exchanger, the working fluid absorbs heat from the waste heat of the flue gas and becomes a high-temperature, high-pressure gas. This gas enters the screw expander I for isentropic expansion, driving the screw expander I to rotate. After exiting the screw expander I, the high-temperature, high-pressure gas becomes a low-pressure gas and enters the plate heat exchanger for condensation, becoming a subcooled liquid. At the same time, the plate heat exchanger transfers heat to the second circulation loop, turning it into a high-pressure gas. The subcooled liquid from the plate heat exchanger enters the liquid storage tank I, thus completing the circulation in the first circulation loop.

[0008] Furthermore, the second circulation loop includes a liquid storage tank II, a working fluid pump II, a scavenging air heat exchanger, a cylinder liner water heat exchanger, a plate heat exchanger, a screw expander II, and a condenser connected in series. The screw expander II is also connected in series with a generator II, a rectifier II, and an inverter II. The condenser is connected to a seawater pump. The working fluid pump II is connected to a motor II, which is driven by a frequency converter II to draw subcooled liquid from the liquid storage tank II and then pump it to the scavenging air heat exchanger for initial heating. After exiting the scavenging air heat exchanger, the working fluid enters... After further heating in the cylinder liner water heat exchanger, the preheated working fluid enters the plate heat exchanger to absorb heat from the working fluid in the first circulation loop, becoming a high-pressure gas. This gas then enters the screw expander II for isentropic expansion, driving the screw expander II to rotate. After exiting the screw expander II, it becomes a low-pressure gas and enters the condenser for condensation, becoming a subcooled liquid. At this time, the heat of the working fluid is transferred to the seawater in the condenser. The subcooled liquid exiting the condenser enters the storage tank II, thus completing the second circulation loop.

[0009] Furthermore, the screw expander I drives the generator I to generate electricity. The generated electricity is rectified into direct current by rectifier I, and then inverted into power frequency alternating current by inverter I and fed into the ship's electrical grid. The screw expander II drives the generator II to generate electricity. The generated electricity is rectified into direct current by rectifier II, and then inverted into power frequency alternating current by inverter II and fed into the ship's electrical grid.

[0010] Furthermore, the control device includes a data acquisition module, a main unit power and waste heat source temperature prediction module, an intelligent decision-making module, and a low-level control module. The data acquisition module is used to collect environmental data, main unit power, heat source, and operating data of the power generation unit. The main unit power and waste heat source temperature prediction module is used to analyze the collected data to predict the main unit power and waste heat source temperature. The intelligent decision-making module is used to obtain the optimal evaporation pressure and superheat by looking up a table based on the predicted heat source temperature. The low-level control module controls the output of corresponding working fluid pump and expander speed control signals through a nonlinear model predictive control algorithm, so that the evaporator outlet pressure and superheat in the power generation unit reach the set values, thereby maximizing the system output power.

[0011] Furthermore, the data acquisition module includes environmental data acquisition, power generation unit data acquisition, and main engine power and heat source temperature data acquisition. Environmental data acquisition is achieved by installing wave sensors and anemometers on the ship to collect real-time data on wave height, water flow velocity and direction, wind speed and direction. Power generation unit data acquisition is used to collect the pressure and temperature at the evaporator outlet in each loop of the power generation unit, as well as the speed of the expander in the power generation unit. Main engine power and heat source temperature data acquisition is used to collect the power of the main engine, the temperature of flue gas, cylinder liner water, and scavenging air.

[0012] Furthermore, the main engine power and waste heat source temperature prediction module obtains environmental data and power generation unit operation data through communication with the data acquisition module, and predicts the ship's main engine power and multiple waste heat source temperatures; the main engine power and waste heat source temperature prediction module includes a main engine power prediction algorithm, a model continuous learning algorithm, and a waste heat source temperature prediction algorithm.

[0013] Furthermore, in the intelligent decision-making module's operational process, a dynamic mechanism model encompassing all equipment is first established. The core input parameters of this model are the heat source temperature and the cold source temperature, while the variables to be optimized are the evaporation pressure and superheat. Then, the core output parameter of the dynamic mechanism model is the system's net output power. Given the heat source and cold source temperatures, the evaporation pressure and superheat are continuously scanned to create multiple parameter combinations. For each parameter combination, it is input into the dynamic mechanism model, and combined with the current heat source and cold source temperatures, the corresponding system net output power is simulated and calculated. All simulation results are compared, and the evaporation pressure and superheat parameter combination corresponding to the maximum net output power is selected. For different heat source and cold source temperature ranges, a mapping table of "heat source temperature - cold source temperature - optimal evaporation pressure - optimal superheat" is established to enable rapid retrieval of optimal parameters across the entire operating range.

[0014] Furthermore, after receiving the optimal evaporation pressure and optimal superheat reference values, the underlying control module uses a nonlinear model predictive control algorithm to output the corresponding working fluid pump speed and expander speed signals. By controlling the speeds of the working fluid pump and expander, the evaporation pressure and superheat of the first and second circulation loops in the excess heat source coupled power generation device are changed, so that the evaporation pressure and superheat reach the optimal values. The working fluid pump speed signal is sent to the frequency converter used to drive the working fluid pump via RS485 communication. The frequency converter controls the motor speed, thereby changing the speed of the working fluid pump. The expander speed signal is sent to the rectifier via RS485 communication. The rectifier controls the speed of the generator and the expander connected to the generator via a rigid coupling by controlling the output current.

[0015] Compared with the prior art, the present invention has the following advantages: The power generation system of this invention can simultaneously recover waste heat from ship scavenging air, cylinder liner water, and flue gas, and improve energy utilization efficiency through dual-circulation loop coupling; and through intelligent control device, it can predict the main engine power based on sea wave conditions, water flow speed, and wind speed, and then predict the recoverable heat, so as to achieve precise control of the power generation device and keep it in a high-efficiency operating state. Attached Figure Description

[0016] Figure 1 This is a structural block diagram of the power generation system of the present invention; Figure 2This is a schematic diagram of the power generation system of the present invention; Figure 3 This is a block diagram of the control device of the present invention; wherein, ① is the outlet temperature of the first circulating evaporator, ② is the outlet pressure of the first circulating evaporator, ③ is the rotational speed of the first circulating expander, ④ is the outlet temperature of the second circulating evaporator, ⑤ is the outlet pressure of the second circulating evaporator, ⑥ is the rotational speed of the second circulating expander, ⑦ is the speed control signal of the first circulating working fluid pump, ⑧ is the speed control signal of the second circulating working fluid pump, ⑨ is the speed control signal of the first circulating expander, and ⑩ is the speed control signal of the second circulating expander. Figure 4 This is a flowchart of the control device of the present invention.

[0017] In the diagram: 1-Storage tank I, 2-Working fluid pump I, 3-Shell-tube heat exchanger, 4-Screw expander I, 5-Plate heat exchanger, 6-Cylinder liner water heat exchanger, 7-Scavenging air heat exchanger, 8-Working fluid pump II, 9-Storage tank II, 10-Condenser, 11-Screw expander II, 12-Generator I, 13-Generator II, 14-Rectifier I, 15-Inverter I, 16-Rectifier II, 17-Inverter II, 18-Seawater pump, 19-Motor I, 20-Variable frequency drive I, 21-Motor II, 22-Variable frequency drive II, 23-Seawater pump motor, 24-Control device. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] like Figure 1 and Figure 2 As shown, a ship heat source coupled power generation system includes a waste heat power generation device and a control device. The waste heat power generation device includes a first circulation loop and a second circulation loop. The evaporation pressure and superheat in the first and second circulation loops are controlled by the control device, enabling simultaneous power generation using three types of waste heat: scavenging air, cylinder liner water, and flue gas. The control device collects wave and wind speed data, uses a convolutional neural network to predict the main engine power and heat source temperature, outputs the corresponding optimal evaporation pressure and optimal superheat based on the prediction, and controls the evaporation pressure and superheat in the first and second circulation loops by using a nonlinear model prediction algorithm to control the speed of the expander and the working fluid pump in the first and second circulation loops, thereby maximizing the output power of the power generation device and improving the waste heat source utilization efficiency.

[0020] The power generation system of this invention comprises two circulation loops, capable of simultaneously generating electricity using three types of waste heat: scavenging air, cylinder liner water, and flue gas. Both the first and second circulation loops are based on organic Rankine cycle technology. The first circulation loop recovers waste heat from the ship's main engine flue gas, generating electricity from the flue gas; the second circulation loop recovers waste heat from the scavenging air and cylinder liner water, generating electricity from the cylinder liner water and scavenging air. The first and second circulation loops are connected to a plate heat exchanger 5, and the electricity output from both loops is fed into the ship's power grid via an inverter. Each evaporator outlet in the power generation system is equipped with a temperature sensor and a pressure sensor. An encoder is mounted on the expander shaft to measure the expander speed, and the data is transmitted to the control device via bus communication. The speed of the working fluid pump in the waste heat source power generation device is controlled by a frequency converter-driven motor, and the expander speed is controlled by the rectifier controlling the current. The control signal is sent by the control device to the frequency converter and rectifier in the waste heat source power generation device via bus communication.

[0021] Specifically, the first circulation loop includes a liquid storage tank I1, a working fluid pump I2, a shell-and-tube heat exchanger 3, a screw expander I4, and a plate heat exchanger 5 connected in series. The screw expander I4 is also connected in series with a generator I12, a rectifier I14, and an inverter I15. The working fluid pump I2 is connected to a motor I19, which is driven by a frequency converter I20 to draw subcooled liquid from the liquid storage tank I1 and then pump it into the shell-and-tube heat exchanger 3. The working fluid absorbs heat from the waste heat of the flue gas in the shell-and-tube heat exchanger 3. The gas becomes a high-temperature, high-pressure gas, which enters the screw expander I4 for isentropic expansion, driving the screw expander I4 to rotate. After exiting the screw expander I4, the high-temperature, high-pressure gas becomes a low-pressure gas, which enters the plate heat exchanger 5 for condensation, becoming a subcooled liquid. At the same time, the plate heat exchanger 5 transfers heat to the second circulation loop, turning it into a high-pressure gas. The subcooled liquid exiting the plate heat exchanger 5 enters the storage tank I1, thus completing the cycle in the first circulation loop.

[0022] Furthermore, the second circulation loop includes a liquid storage tank II9, a working fluid pump II8, a scavenging air heat exchanger 7, a cylinder liner water heat exchanger 6, a plate heat exchanger 5, a screw expander II11, and a condenser 10 connected in series. The screw expander II11 is also connected in series with a generator II13, a rectifier II16, and an inverter II17. The condenser 10 is connected to a seawater pump 18. The working fluid pump II8 is connected to a motor II21, which is driven by a frequency converter II22 to draw subcooled liquid from the liquid storage tank II9 and then pump it to the scavenging air heat exchanger 7 for initial heating. The working fluid is drawn from the scavenging air heat exchanger... After exiting from 7, the working fluid enters the cylinder liner water heat exchanger 6 for further heating. The preheated working fluid then enters the plate heat exchanger 5 to absorb heat from the working fluid in the first circulation loop, becoming high-pressure gas. This gas then enters the screw expander II 11 for isentropic expansion, driving the screw expander II 11 to rotate. After exiting the screw expander II 11, the gas becomes low-pressure gas and enters the condenser 10 for condensation, becoming subcooled liquid. At this time, the heat of the working fluid is transferred to the seawater in the condenser 10. The subcooled liquid exiting the condenser 10 enters the storage tank II 9, thus completing the second circulation loop.

[0023] Screw expander I4 is rigidly connected to generator I12, and screw expander II11 is rigidly connected to generator II13 via couplings. This ensures that the mechanical energy output from screw expander I4 and screw expander II11 can be efficiently transferred to generator I12 and generator II13, respectively, converting mechanical energy into electrical energy. To achieve high system efficiency, the speeds of generator I12 and generator II13 need to be controlled. Therefore, the downstream ends of generator I12 and generator II13 in the two circulation loops are connected to rectifier I14 and rectifier II16, respectively. The speeds of the generators and their connected expanders can be controlled through the rectifiers, and the electricity output from the rectifiers is fed into the ship's electrical grid through an inverter. Specifically: the screw expander I4 drives the generator I12 to generate electricity. The generated electricity is rectified into direct current by rectifier I14, and then inverted into power frequency alternating current by inverter I15 and fed into the ship's power grid; the screw expander II11 drives the generator II13 to generate electricity. The generated electricity is rectified into direct current by rectifier II16, and then inverted into power frequency alternating current by inverter II17 and fed into the ship's power grid.

[0024] In the power generation system of this application, the liquid storage tank I1 is used to store condensate from the condenser as a buffer; the working fluid pump I2 provides circulation power for the working fluid flow in the first circulation loop; the shell-and-tube heat exchanger 3 is used to exchange heat with the waste heat of the flue gas, and simultaneously heat the working fluid to make it high-pressure, high-temperature steam; the screw expander I4 converts the energy carried in the high-temperature, high-pressure steam into mechanical energy output; the plate heat exchanger 5 condenses the working fluid in the first circulation loop into a liquid, and simultaneously heats the working fluid in the second circulation loop to make it high-pressure, high-temperature steam; the cylinder liner water heat exchanger 6 is used to absorb heat from the cylinder liner water, and simultaneously... The working fluid in the second circulation loop is heated; the scavenging air heat exchanger 7 absorbs heat from the scavenging air and simultaneously heats the working fluid in the second circulation loop; the working fluid pump II 8 provides circulation power for the working fluid flow in the second circulation loop; the liquid storage tank II 9 stores condensed liquid from the condenser as a buffer; the condenser 10 condenses the gaseous circulating working fluid into a liquid; the screw expander II 11 converts the energy carried in the high-temperature, high-pressure steam into mechanical energy output; the generator I 12 converts mechanical energy into electrical energy; the generator II 13 converts mechanical energy into electrical energy; and the rectifier I 14 converts the generated energy into electrical energy. The alternating current generated by generator I12 is converted into direct current, and the speed of screw expander I4 can be controlled by controlling the output current. Inverter I15 converts the direct current into AC power at the industrial frequency, which can be connected to the ship's electrical grid. Rectifier II16 converts the alternating current generated by generator II13 into direct current, and the speed of screw expander II11 can be controlled by controlling the output current. Inverter II17 converts the direct current into AC power at the industrial frequency, which can be connected to the ship's electrical grid. Seawater pump 18 draws seawater and sends it to condenser 10 in the second circulation loop. Motor I19 provides power to working fluid pump I2, driving it to rotate. Variable frequency drive I2... Motor I19 is driven by inverter I20, which controls the speed of motor I19 and thus the speed of working fluid pump I2. Motor II21 provides power to working fluid pump II8, driving it to rotate. Inverter II22 drives motor II21, which controls the speed of motor II21 and thus the speed of working fluid pump II8. Seawater pump motor 23 provides power to seawater pump 18. Control device 24 controls the evaporation pressure and superheat in the system by controlling the speeds of working fluid pumps I2 and II8 and screw expanders I4 and II11 in the waste heat power generation device.

[0025] Continue reading Figure 2It can be seen that the liquid storage tank I1 is connected to the working fluid pump I2 and the shell-and-tube heat exchanger 3 via copper pipes; the screw expander I4 is connected to the shell-and-tube heat exchanger 3 and the plate heat exchanger 5 via high-pressure copper pipes. The cylinder liner water heat exchanger 6 is connected to the scavenging air heat exchanger 7 and the plate heat exchanger 5 via copper pipes; the working fluid pump II8 is connected to the scavenging air heat exchanger 7 and the liquid storage tank II9 via copper pipes; the screw expander II11 is connected to the condenser 10 and the plate heat exchanger 5 via copper pipes. The screw expander I4 is connected to the generator I12 via a rigid coupling 201; the generator I12 is connected to the rectifier I14 via a three-phase cable 203; the rectifier I14 and the inverter I15 are connected via a DC cable; the inverter I15 is connected to the ship's power grid 209 via a busbar.

[0026] Screw expander II11 is connected to generator II13 via a rigid coupling. Generator II13 is connected to rectifier II16 via a three-phase cable 204. Rectifier II16 and inverter II17 are connected via a DC cable, and inverter II17 is connected to the ship's power grid 209 via a busbar. The speed inverter control signals 401 and 402 for working fluid pump I2, 403 and 404 for screw expander I4, 412 and 413 for evaporator outlet pressure in the first circulation loop, 414 and 415 for evaporator outlet pressure in the second circulation loop, 415 and 416 for speed of screw expander I4, and 411 for speed of screw expander II11 are all communicated via RS485 communication and control device 24.

[0027] In one specific embodiment, in the first circulation loop, the flue gas of the shell-and-tube heat exchanger 3 flows through the shell side, while the organic working fluid flows through the tube side. To improve heat exchange efficiency, the heat exchange tubes are made of copper. The circulating working fluid is R245fa, which has good thermophysical properties and environmental characteristics. The working fluid enters the screw expander I4 from the shell-and-tube heat exchanger 3 through pipe 102. The isentropic efficiency of the screw expander I4 is over 80%. The working fluid exiting the screw expander I4 enters the plate heat exchanger 5 through pipe 103. The plate heat exchanger 5 uses a plate heat exchanger with a higher heat transfer coefficient. The working fluid exiting the plate heat exchanger 5 enters the storage tank I1 through pipe 105. The working fluid pump I2 draws the working fluid from the storage tank I1 through pipe 105. The working fluid pump I2 is a corrosion-resistant centrifugal pump. The working fluid pump I2 is driven by the motor driven by the frequency converter I20. The speed control signal 401 of the motor I19 driven by the frequency converter I20 is sent to the communication interface of the frequency converter I20 by the control device 24 through RS485 communication.

[0028] In the second circulation loop, the screw expander II11 also uses a screw-type expander, and the circulating working fluid is R134a. The condenser 10 is a corrosion-resistant plate condenser, using seawater to cool the working fluid. The seawater is pumped into the condenser 10 via the seawater pump 18. The working fluid pump II8, like the working fluid pump I2, is a corrosion-resistant centrifugal pump, driven by the frequency converter II22 to operate the motor II21. The speed control signal 402 of the frequency converter II22 driving the motor II21 is sent to the communication interface of the frequency converter II22 via RS485 communication from the intelligent control device.

[0029] Screw expander I4 and generator I12, and screw expander II11 and generator II13 are connected via rigid couplings 201 and 202, respectively. Generators I12 and II13 are permanent magnet synchronous generators, characterized by their small size, light weight, and high efficiency, making them suitable for use in the limited space of a ship. The control algorithm for rectifiers I14 and II16 employs a zero d-axis current control strategy, and the modulation strategy uses a space voltage vector PWM modulation algorithm. The core device is an IGBT. Inverters I15 and II17 have both off-grid and grid-connected modes. Grid-connected mode uses a grid voltage-oriented vector control strategy, while off-grid mode uses a V / F control algorithm; both also employ a PWM modulation algorithm, and the device is an IGBT. The excess heat power generation unit is equipped with temperature and pressure sensors at the evaporator outlet in the first and second circulation loops. Encoders are installed on the shafts of screw expanders I4 and II11 to measure rotational speed. The rotational speed, temperature, and pressure signals are transmitted to the control device 24 via an RS485 fieldbus.

[0030] The output power of a heat source coupled with a heat source is affected by the temperatures of the heat source and the cold source. To ensure the power generation device outputs maximum power and operates within safe boundaries under given heat source and cold source temperatures, a corresponding control device 24 is required. For example... Figure 3 and Figure 4 As shown, the control device 24 includes a data acquisition module, a main unit power and waste heat source temperature prediction module, an intelligent decision-making module, and a low-level control module. The data acquisition module is used to collect environmental data, main unit power, heat source, and operating data of the power generation device. The main unit power and waste heat source temperature prediction module is used to analyze the collected data to predict the main unit power and waste heat source temperature. The intelligent decision-making module is used to obtain the optimal evaporation pressure and superheat by looking up a table based on the predicted heat source temperature. The low-level control module controls the output of corresponding working fluid pump and expander speed control signals through a nonlinear model predictive control algorithm, so that the evaporator outlet pressure and superheat in the power generation device reach the set values, thereby maximizing the system output power.

[0031] (1) The data acquisition module includes environmental data acquisition, power generation device data acquisition, and main engine power and heat source temperature data acquisition. Environmental data acquisition is carried out by installing wave sensors and anemometers on the ship to collect data on wave height, water flow velocity and direction, wind speed and wind direction in real time. Power generation device data acquisition is used to collect the pressure and temperature of the evaporator outlet in each loop of the power generation device and the speed of the expander in the power generation device. Main engine power and heat source temperature data acquisition is used to collect the power of the main engine, flue gas, cylinder liner water and scavenging air temperature.

[0032] In one specific embodiment, the data acquisition module uses TI's SN75176 as the RS485 communication module, and the data acquisition processing and data transmission and reception control use the STM32F103C8T6 chip.

[0033] (2) The main engine power and waste heat source temperature prediction module obtains environmental data and power generation device operation data through communication with the data acquisition module, and predicts the main engine power and multiple waste heat source temperatures of the ship; the main engine power and waste heat source temperature prediction module includes a main engine power prediction algorithm, a model continuous learning algorithm, and a waste heat source temperature prediction algorithm.

[0034] 1) The specific process of the host power prediction algorithm is as follows: ① Data acquisition and preprocessing Data collection scope: Historical data includes wave parameters (wave height, period, wave direction, current speed), wind speed parameters (wind speed, wind direction), and main engine power during ship navigation, with a sample size of 200,000 records and a time granularity of 1 second / record.

[0035] The preprocessing steps include outlier removal, normalization, and feature reconstruction. Outlier removal uses the 3σ criterion to eliminate outliers in wave, wind speed, and power data. During normalization, the input features are normalized to the [0,1] interval using min-max normalization. In feature reconstruction, the time-series data is converted into a two-dimensional feature matrix, with parameters from 10 consecutive seconds forming a 10×6 feature matrix (10 time steps, 6 features), which serves as the input sample for the convolutional neural network.

[0036] ② Convolutional Neural Network (CNN) Model Design The CNN network consists of one input layer, two convolutional layers, two pooling layers, one fully connected layer, and an output layer. The input layer receives a 10×6 feature matrix; Convolutional Layer 1 uses 16 3×3 convolutional kernels with a stride of 1, same padding, and ReLU activation, producing an output feature map of 10×6×16; Pooling Layer 1 uses 2×2 max pooling with a stride of 2, producing an output feature map of 6×3×16; Convolutional Layer 2 uses 32 2×2 convolutional kernels with a stride of 1, valid padding, and ReLU activation, producing an output feature map of 4×2×32; Pooling Layer 2 uses 2×2 average pooling with a stride of 1, producing an output feature map of 3×1×32; The output layer has one node (linear activation) and outputs the predicted host power.

[0037] ③ Model loss The loss function is the root mean square error (RMSE), the optimizer is the Adam optimizer, the initial learning rate is 0.001, and it decays by 10% every 50 rounds. The training strategy is to use 5-fold cross-validation, with 200 training rounds and an early stopping strategy (stop if the MSE on the validation set does not decrease for 10 consecutive rounds).

[0038] ④ Model Training The wave height, period, wave direction, and flow velocity collected by the wave sensor, and the wind and wind direction collected by the anemometer are preprocessed to construct a 6-dimensional feature vector at time t. The feature vectors of the most recent 10 seconds are taken by a sliding window to form a 10×6 real-time feature matrix. The trained CNN model is input to output the predicted host power value at time t.

[0039] 2) The specific process of the model continuous learning algorithm is as follows: ① Online learning data input The system collects real-time data from the host power sensor (actual power) and heat source temperature sensor data (flue gas temperature, cylinder liner water temperature, and scavenging air temperature), generating one sample every 30 seconds. It calculates the error between the predicted power and the actual power, and marks the sample as a "high-value sample" when the error exceeds a preset threshold.

[0040] ② Incremental training strategy First, data weights are assigned: high-value samples are weighted at 1.5, ordinary samples at 1.0, and historical samples have their weights decayed over time. Then, incremental updates are performed, with incremental training triggered every 24 hours. The parameters of the first two layers of the CNN are frozen, and only the fully connected layers are updated. Mini-batch gradient descent is used with a learning rate of 1 / 10 of the initial value to avoid catastrophic forgetting. Finally, the model is evaluated by calculating the MSE using a rolling validation set (data from the last 7 days). When the MSE decreases by ≥3%, the updated model is saved.

[0041] 3) The specific process of the waste heat source temperature prediction algorithm is as follows: The prediction of waste heat source temperature is based on the first law of thermodynamics and is calculated using a formula that fits the temperature of the main engine power. The formula for predicting flue gas temperature is as follows: , in, T exh For flue gas temperature, P t The parameters are: main unit power, a1, b1, and c1. The parameters are obtained by fitting historical power data and measured flue gas temperature using the least squares method.

[0042] The formula for predicting cylinder liner water temperature is as follows: , in, T jac For cylinder liner water temperature, P t The parameters a2, b2, and c2 are the main engine power and the fitting parameters, which are obtained by fitting historical power data with measured cylinder liner water temperature using the least squares method.

[0043] The scavenging air temperature prediction is calculated using the heat exchange calculation formula: , in, T scav The temperature of the scavenging air. K The heat transfer coefficient, A For heat exchange area, ΔT The average heat exchange temperature difference is given by c, where c is the specific heat capacity of the scavenged air. m This is the scavenging airflow rate. T in The inlet temperature of the scavenging air is usually taken as the ambient temperature.

[0044] In one specific embodiment, the main engine power and heat source temperature prediction module includes a core controller and multiple sensors. The core controller uses an NVIDIA Jetson TX2 NX with a computing power of 1.33 TFLOPS. The wave sensor uses an acoustic Doppler wave velocity profiler, installed below the bow waterline, capable of accurately measuring parameters such as wave height, period, and water flow velocity. The wind speed sensor uses a high-precision marine anemometer with corrosion resistance. The mathematical model uses a convolutional neural network model, with input data including wave period, wave height, water flow velocity, wind speed, wind direction, and heading, and output parameter being the main engine power. The training data exceeds 200,000 sets to ensure model accuracy. The heat source temperature prediction module establishes a relationship model corresponding to multiple heat source temperature parameters based on the main engine power. Through experimental testing, the main engine power and the heat source temperatures of scavenging air, cylinder liner water, and flue gas are obtained, and linear fitting is performed, with an exponent of 5. The main engine power and heat source temperature prediction module communicates with the intelligent decision-making module via RS485.

[0045] (3) In the intelligent decision-making module, a dynamic mechanism model containing all equipment is first established. The core input parameters of the dynamic mechanism model are the heat source temperature and the cold source temperature, and the variables to be optimized are the evaporation pressure and the superheat. The core output parameter of the dynamic mechanism model is the net output power of the system. Under the given heat source temperature and cold source temperature, the evaporation pressure and the superheat are changed by continuous scanning to form multiple sets of parameter combinations. For each set of parameters, it is input into the dynamic mechanism model, and the corresponding net output power of the system is calculated by simulation in combination with the current heat source temperature and cold source temperature. All simulation results are compared, and the evaporation pressure and superheat parameter combination corresponding to the maximum value of the net output power is selected. For different heat source temperature ranges and cold source temperature ranges, the above steps are repeated to establish a mapping relationship table of "heat source temperature - cold source temperature - optimal evaporation pressure - optimal superheat" to realize the rapid call of the optimal parameters in the entire operating range.

[0046] In one specific embodiment, the intelligent decision-making module uses an industrial-grade TMS320F280039C as its core controller. It simulates the optimal evaporation pressure and optimal superheat corresponding to maximum power output under different heat source temperatures by establishing a mechanistic model. The data between the optimal evaporation pressure and optimal superheat and the corresponding heat source temperature are fitted and stored in the controller. Upon receiving predicted waste heat source temperature data via RS485 communication, the module outputs reference values ​​for the optimal evaporation pressure and optimal superheat using the fitted formula. The intelligent decision-making module and the underlying control module communicate via RS485.

[0047] (4) After receiving the optimal evaporation pressure and optimal superheat reference values, the bottom control module uses a nonlinear model predictive control algorithm to output the corresponding working fluid pump speed and expander speed signals. By controlling the working fluid pump and expander speeds, the evaporation pressure and superheat of the first and second circulation loops in the excess heat source coupled power generation device are changed, so that the evaporation pressure and superheat reach the optimal values. The working fluid pump speed signal is sent to the frequency converter used to drive the working fluid pump through RS485 communication. The frequency converter controls the motor speed, thereby changing the working fluid pump speed. The expander speed signal is sent to the rectifier through RS485 communication. The rectifier controls the speed of the generator and the expander connected to the generator through a rigid coupling by controlling the output current.

[0048] The underlying controller algorithm employs a nonlinear model predictive control (NMPC) algorithm. By integrating the nonlinear dynamic characteristics of the power generation unit, multivariable constraints, and a rolling optimization mechanism, it achieves high-precision control of the optimal evaporation pressure and optimal superheat. By setting boundary conditions, the system can operate within safe boundaries. The control principle is based on real-time measurement information, solving a constrained nonlinear optimization problem to dynamically generate control sequences for the working fluid pump and expander speeds. The specific process is as follows: ① Initialization and State Awareness (Sampling Time) Based on real-time data acquisition from sensors of current status (evaporation pressure and superheat), combined with feedback from the working fluid pump and expander speed, and by using an extended Kalman filter (EKF) to eliminate measurement noise, an accurate initial state is obtained.

[0049] ② Rolling optimization (solving within the prediction time domain) Starting from the initial state, the state sequence for the next N steps is predicted based on a nonlinear dynamic model. Under the premise of satisfying safety constraints (state, control variables and derived constraints), the control sequence that minimizes the objective function is solved in the control time domain.

[0050] The optimization problem is solved using the interior point method. By transforming the inequality constraints into barrier functions and combining them with the gradient descent algorithm, the problem converges quickly, ensuring that the calculation is completed within the sampling period.

[0051] ③ Control quantity execution The first control variable of the optimized sequence is executed, and the working fluid pump and expander speed signals are sent to the corresponding actuators to achieve real-time control of evaporation pressure and superheat.

[0052] ④ Feedback correction (proceeding to the next sampling time) New state measurements are collected, prediction errors are calculated, and the initial state of the prediction model in the next round is corrected through a state error compensation mechanism to offset the effects of model mismatch and external disturbances and ensure control robustness.

[0053] In one specific embodiment, the underlying control module uses a TMS320F28P550SJ as the core controller, which has hardware calculation units for floating-point numbers and trigonometric functions, resulting in high real-time control performance. The controller runs a model predictive control algorithm, using the open-source ACADO Codegen solver. Communication between the controller, inverter, and rectifier is via RS485. The underlying control module controls the speeds of the working fluid pump and expander by outputting control signals to the working fluid pump inverter and expander speed control signals to the rectifier. By controlling the working fluid pump and expander speeds, the evaporation pressure and superheat in each cycle of the power generation unit are controlled.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A ship heat source coupled power generation system, characterized in that, The device includes a waste heat power generation unit and a control unit. The waste heat power generation unit comprises a first circulation loop and a second circulation loop. The evaporation pressure and superheat in the first and second circulation loops are controlled by the control unit, enabling simultaneous power generation using three types of waste heat: scavenging air, cylinder liner water, and flue gas. The control unit collects wave and wind speed data, uses a convolutional neural network to predict the main unit power and heat source temperature, outputs the corresponding optimal evaporation pressure and optimal superheat based on the prediction, and controls the expander speed and working fluid pump speed in the first and second circulation loops through a nonlinear model prediction algorithm to control the evaporation pressure and superheat in the first and second circulation loops, thereby maximizing the output power of the power generation unit.

2. The ship heat source coupled power generation system according to claim 1, characterized in that, Both the first and second circulation loops generate electricity based on organic Rankine cycle technology. The first circulation loop is used to recover waste heat from the flue gas of the ship's main engine and generate electricity using the flue gas. The second circulation loop is used to recover waste heat from the scavenging air and cylinder liner water and generate electricity using the cylinder liner water and scavenging air. The first and second circulation loops are connected to a plate heat exchanger (5). The electricity output from the two circulation loops is connected to the ship's power grid through an inverter. Each evaporator outlet of the power generation system is equipped with a temperature sensor and a pressure sensor. An encoder is installed on the expander shaft to measure the expander speed and transmits the data to the control device through bus communication. The speed of the working fluid pump in the waste heat source power generation device is controlled by a frequency converter-driven motor. The speed of the expander is controlled by the rectifier to control the current. The control signal is sent by the control device to the frequency converter and rectifier in the waste heat source power generation device through bus communication.

3. A ship heat source coupled power generation system according to claim 2, characterized in that, The first circulation loop includes a liquid storage tank I (1), a working fluid pump I (2), a shell-and-tube heat exchanger (3), a screw expander I (4), and a plate heat exchanger (5) connected in series. The screw expander I (4) is also connected in series with a generator I (12), a rectifier I (14), and an inverter I (15). The working fluid pump I (2) is connected to a motor I (19), which is driven by a frequency converter I (20) to drive the working fluid pump I (2) to draw subcooled liquid from the liquid storage tank I (1) and then pump it into the shell-and-tube heat exchanger (3). The working fluid in the shell-and-tube heat exchanger (3) 3) The heat absorbed by the flue gas is transformed into high-temperature and high-pressure gas. This gas will enter the screw expander I (4) for isentropic expansion, driving the screw expander I (4) to rotate. After the high-temperature and high-pressure gas comes out of the screw expander I (4), it becomes low-pressure gas. The low-pressure gas enters the plate heat exchanger (5) for condensation and becomes subcooled liquid. At the same time, the plate heat exchanger (5) will transfer heat to the second circulation loop, making it into high-pressure gas. The subcooled liquid from the plate heat exchanger (5) enters the storage tank I (1), thus completing the circulation in the first circulation loop.

4. A ship heat source coupled power generation system according to claim 3, characterized in that, The second circulation loop includes a liquid storage tank II (9), a working fluid pump II (8), a scavenging air heat exchanger (7), a cylinder liner water heat exchanger (6), a plate heat exchanger (5), a screw expander II (11), and a condenser (10) connected in series. The screw expander II (11) is also connected in series with a generator II (13), a rectifier II (16), and an inverter II (17). The condenser (10) is connected to a seawater pump (18). The working fluid pump II (8) is connected to a motor II (21). The motor II (21) is driven by a frequency converter II (22) to drive the working fluid pump II (8) to draw subcooled liquid from the liquid storage tank II (9) and then pump it to the scavenging air heat exchanger (7) for the first time. Heating; After the working fluid comes out of the scavenging air heat exchanger (7), it enters the cylinder liner water heat exchanger (6) for further heating. The preheated working fluid then enters the plate heat exchanger (5) to absorb the heat in the working fluid of the first circulation loop and becomes high-pressure gas. This gas enters the screw expander II (11) for isentropic expansion, driving the screw expander II (11) to rotate. After coming out of the screw expander II (11), it becomes low-pressure gas and enters the condenser (10) for condensation, becoming subcooled liquid. At this time, the heat of the working fluid is transferred to the seawater in the condenser (10). The subcooled liquid coming out of the condenser (10) enters the storage tank II (9), thus completing the second circulation loop.

5. A ship heat source coupled power generation system according to claim 4, characterized in that, The screw expander I (4) drives the generator I (12) to generate electricity. The generated electricity is rectified into DC by rectifier I (14) and then inverted into AC by inverter I (15) and fed into the ship's power grid. The screw expander II (11) drives the generator II (13) to generate electricity. The generated electricity is rectified into DC by rectifier II (16) and then inverted into AC by inverter II (17) and fed into the ship's power grid.

6. A ship heat source coupled power generation system according to claim 4, characterized in that, The control device (24) includes a data acquisition module, a main unit power and waste heat source temperature prediction module, an intelligent decision-making module, and a low-level control module. The data acquisition module is used to collect environmental data, main unit power, heat source, and operating data of the power generation device. The main unit power and waste heat source temperature prediction module is used to predict the main unit power and waste heat source temperature by analyzing the collected data. The intelligent decision-making module is used to obtain the optimal evaporation pressure and superheat by looking up a table based on the predicted heat source temperature. The low-level control module controls the output of the corresponding working fluid pump and expander speed control signals through a nonlinear model predictive control algorithm, so that the evaporator outlet pressure and superheat in the power generation device reach the set value, thereby maximizing the system output power.

7. A ship heat source coupled power generation system according to claim 6, characterized in that, The data acquisition module includes environmental data acquisition, power generation device data acquisition, and main engine power and heat source temperature data acquisition. The environmental data acquisition is achieved by installing wave sensors and anemometers on the ship to collect real-time data on wave height, water flow velocity and direction, wind speed and direction. The power generation device data acquisition is used to collect the pressure and temperature at the evaporator outlet in each loop of the power generation device, and the speed of the expander in the power generation device. The main engine power and heat source temperature data acquisition is used to collect the power of the main engine, the temperature of flue gas, cylinder liner water, and scavenging air.

8. A ship heat source coupled power generation system according to claim 6, characterized in that, The main engine power and waste heat source temperature prediction module obtains environmental data and power generation device operation data through communication with the data acquisition module, and predicts the ship's main engine power and multiple waste heat source temperatures; the main engine power and waste heat source temperature prediction module includes a main engine power prediction algorithm, a model continuous learning algorithm, and a waste heat source temperature prediction algorithm.

9. A ship heat source coupled power generation system according to claim 6, characterized in that, In the operation of the intelligent decision-making module, a dynamic mechanism model encompassing all equipment is first established. The core input parameters of the dynamic mechanism model are the heat source temperature and the cold source temperature, while the variables to be optimized are the evaporation pressure and superheat. Then, the core output parameter of the dynamic mechanism model is the system's net output power. Given the heat source and cold source temperatures, the evaporation pressure and superheat are continuously scanned to create multiple parameter combinations. For each parameter combination, it is input into the dynamic mechanism model, and combined with the current heat source and cold source temperatures, the corresponding system net output power is simulated and calculated. All simulation results are compared, and the evaporation pressure and superheat parameter combination corresponding to the maximum net output power is selected. For different heat source and cold source temperature ranges, a mapping table of "heat source temperature - cold source temperature - optimal evaporation pressure - optimal superheat" is established to enable rapid retrieval of optimal parameters across the entire operating range.

10. A ship heat source coupled power generation system according to claim 6, characterized in that, After receiving the optimal evaporation pressure and optimal superheat reference values, the underlying control module uses a nonlinear model predictive control algorithm to output corresponding working fluid pump speed and expander speed signals. By controlling the speeds of the working fluid pump and expander, the evaporation pressure and superheat of the first and second circulation loops in the excess heat source coupled power generation device are changed, so that the evaporation pressure and superheat reach the optimal values. The working fluid pump speed signal is sent to the frequency converter used to drive the working fluid pump via RS485 communication. The frequency converter controls the motor speed, thereby changing the speed of the working fluid pump. The expander speed signal is sent to the rectifier via RS485 communication. The rectifier controls the output current to control the speed of the generator and the expander connected to the generator via a rigid coupling.

Images