Energy cascade utilization method and system based on multi-level embedded control

Abstract

The present invention provides a method and system for cascade utilization of energy based on multi-level embedded control, comprising the following steps: step S1: real-time monitoring of the temperature T, flow rate Q and composition data of flue gas discharged by production equipment, and establishing a dynamic distribution function of flue gas energy resources discharged; step S2: dividing the three-level recovery stage based on the temperature threshold interval, and determining the start-stop combination of heat exchange equipment in each stage through a gradient decision model; step S3: constructing a multi-level energy efficiency optimization function and using a neural network to predict the change trend of flue gas energy parameters in the future period; step S4: using a feedback adjustment module to compensate for the deviation between the actual recovery efficiency and the theoretical value, and updating the weight coefficient of the multi-level energy efficiency optimization function. The present invention establishes a dynamic distribution function by monitoring the temperature, flow rate and composition of the flue gas of the production equipment, and uses a gradient decision model and a three-level temperature interval recovery strategy to achieve efficient energy recovery in the full temperature range using different heat exchange equipment.

Description

Technical Field

[0001] The present invention relates to the field of energy-saving technology, and in particular to a method and system for cascaded energy utilization based on multi-level embedded control. Background Art

[0002] As thermal energy equipment widely used in industrial production, the high-temperature flue gas emitted by production equipment contains a large amount of energy resources. According to statistics, the energy loss of flue gas from industrial production equipment accounts for about 15%-30% of the total calorific value of the fuel. If this part of energy can be effectively recovered, it can significantly improve energy utilization and reduce carbon emissions. Traditional energy recovery technology usually uses a single heat exchange device (such as heat pipe heat exchangers, economizers, etc.) to recover energy from flue gas in a specific temperature range. However, a single device can only operate efficiently within its designed temperature range and cannot cover the full temperature range of the flue gas from production equipment. During the operation of production equipment, the flue gas temperature, flow rate and composition fluctuate frequently due to factors such as fuel type and load rate. The existing fixed threshold control strategy is difficult to adjust the equipment operation status in time, resulting in frequent start and stop of heat exchange equipment or long-term deviation from the optimal operating conditions, which not only reduces system stability but also accelerates equipment loss.

[0003] Therefore, there is an urgent need for an energy cascade utilization method and system that can dynamically sense flue gas parameters, intelligently decide on equipment start and stop combinations, and optimize energy efficiency and economy in real time, so as to achieve efficient, stable, and low-cost energy recovery over the entire temperature range. To address the above problems, a method and system for energy cascade utilization based on multi-level embedded control is now being developed. Summary of the invention

[0004] In order to overcome the shortcomings that conventional technology mainly adopts a single heat exchange device to recover flue gas energy within the corresponding temperature range, but a single device can only operate efficiently in a specific temperature range, cannot adapt to the full temperature range of flue gas from production equipment, and low-temperature energy is difficult to utilize, resulting in a low overall heat recovery rate. At the same time, during operation, flue gas parameters fluctuate frequently, and fixed threshold control strategies are prone to cause frequent start and stop of equipment or a sudden drop in efficiency, the present invention provides an energy cascade utilization method and system based on multi-stage embedded control.

[0005] The technical solution of the present invention is: a method for cascade utilization of energy based on multi-level embedded control, comprising the following steps: Step S1: Real-time monitoring of the temperature T, flow rate Q and composition data of flue gas emitted by production equipment, and establishment of a dynamic distribution function f (T, Q, t) of the flue gas energy resources emitted, where t represents the time variable in the flue gas emission process; Step S2: dividing the three-stage recovery phase based on the temperature threshold interval, and determining the start and stop combination of the heat exchange equipment in each phase through the gradient decision model; Step S3: construct a multi-level energy efficiency optimization function and use a neural network to predict the change trend of flue gas energy parameters in the future period, dynamically adjust the classification threshold interval and the operating parameters of the corresponding heat exchange equipment; Step S4: The feedback adjustment module is used to compensate for the deviation between the actual recovery efficiency and the theoretical value, and the weight coefficient of the multi-level energy efficiency optimization function is updated.

[0006] As a preferred embodiment of the present invention, the gradient decision model is constructed in the following manner: Establish the corresponding mapping matrix between relevant temperature and equipment ,when The i-th type of equipment is enabled when , where i=1, 2, 3, corresponding to the heat exchanger, absorption heat pump and ORC power generation device, and j=1, 2, 3, representing the three-level recovery stage divided based on the temperature threshold interval.

[0007] As a preferred embodiment of the present invention, the constraints of the multi-level energy efficiency optimization function include: Constraints on the flue gas dew point temperature threshold: ,in represents the temperature of the flue gas at the jth stage of recovery, Indicates safety margin; Equipment maximum compliance constraint: The actual heat load of the i-th type of equipment needs to be less than the maximum allowable heat load of the corresponding equipment; Economic constraints: ,in is the unit energy consumption cost of the i-th type of equipment, is the operating time of the i-th type of equipment, is the maximum allowable cost, is the operating power of the i-th category equipment.

[0008] As a preferred embodiment of the present invention, the input parameters for neural network training include: historical operation data, ambient temperature and humidity, production equipment load rate, fuel calorific value, and the output parameters are the temperature prediction value and confidence interval for the next 15 minutes.

[0009] An energy cascade utilization system based on multi-level embedded control, characterized by comprising: A multi-stage heat exchange device group, including a high-temperature section heat pipe heat exchanger, a medium-temperature section absorption heat pump, and a low-temperature section ORC power generation device; Dynamic monitoring module, equipped with infrared temperature measurement array and ultrasonic flow meter, collects flue gas parameters in real time; Intelligent control unit, integrated control algorithm to intelligently control the system; Model update module, which stores historical operation data and regularly optimizes neural network model parameters.

[0010] As a preferred embodiment of the present invention, the multi-stage heat exchange device group adopts a series-parallel mixed arrangement, the high temperature section and the medium temperature section are connected in series, the low temperature section is provided with a parallel bypass, and adjustable baffles are provided between each section.

[0011] As a preferred embodiment of the present invention, the intelligent control unit comprises: Decision engine, which executes the gradient decision model to output equipment start and stop instructions; Optimization calculator, solving the extreme values ​​of multi-level energy efficiency optimization functions in real time; The prediction corrector performs Kalman filtering to fuse the output of the neural network with the measured values.

[0012] As a preferred embodiment of the present invention, it also includes a visual human-machine interface, which dynamically displays a three-level recovery efficiency comparison radar chart, a multi-level energy efficiency optimization function convergence curve, and a device operation status topology chart.

[0013] By adopting the above technical solution, the present invention has the following advantages: 1. The present invention monitors the temperature, flow rate and composition data of the flue gas emitted by the production equipment in real time, establishes a dynamic distribution function based on these data, and adopts a gradient decision model to determine the start and stop combination of equipment in different temperature ranges, thereby realizing efficient recovery of the energy of the production equipment. At the same time, three recovery stages are divided based on the temperature threshold range, and different types of heat exchange equipment are used for step-by-step recovery, so that the system can adapt to changes in the full temperature range of the flue gas of the production equipment.

[0014] 2. The present invention predicts the changing trend of flue gas energy parameters in the future period through a neural network, dynamically adjusts the classification threshold interval and the operating parameters of the corresponding equipment, thereby improving the overall heat recovery efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention is a flowchart of the steps of the energy cascade utilization method based on multi-level embedded control.

[0016] Figure 2 It is a structural schematic diagram of the energy cascade utilization system based on multi-level embedded control of the present invention.

[0017] Figure 3 It is a schematic structural diagram of the multi-stage heat exchange device group of the present invention.

[0018] Figure 4 It is a structural schematic diagram of the intelligent control unit of the present invention. DETAILED DESCRIPTION

[0019] Reference to an embodiment herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0020] A method for cascade utilization of energy based on multi-level embedded control, such as Figure 1 As shown, the following steps are included: Step S1: An infrared temperature measurement array (model FLIR A615, temperature range -40°C-650°C, accuracy ±1°C) is installed at the boiler flue outlet to collect temperature data every 10 seconds, and an ultrasonic flow meter (model Siemens SitransFUS380, measurement error ±0.5%) is used to monitor the flue gas volume flow in real time. The temperature T, flow Q and composition data of the flue gas discharged from the boiler are monitored in real time. The composition data is obtained by using a flue gas composition analyzer, model Testo 350, to detect the contents of O2, CO2 and H2O, and to establish a dynamic distribution function f (T, Q, t) of the flue gas waste heat resources discharged, where t represents the time variable in the flue gas emission process; Step S2: Divide the recovery stage into three stages based on the temperature threshold interval. Specifically, the high temperature stage , medium temperature section , low temperature section , determine the start and stop combination of heat exchange equipment in each stage through the gradient decision model; Step S3: Construct a multi-level energy efficiency optimization function and use a neural network to predict the changing trend of flue gas waste heat parameters in the future period. During the prediction process, the historical temperature, flow rate, and boiler load rate (sampling period 1 minute, time window 60 minutes) are input to obtain the output temperature forecast value for the next 15 minutes. and 95% confidence interval, if the predicted temperature If the predicted temperature exceeds the current interval threshold for 5 consecutive minutes If it exceeds the current range, the threshold is adjusted in steps of 10°C; Step S4: The feedback adjustment module is used to compensate for the deviation between the actual recovery efficiency and the theoretical value, and the weight coefficient of the multi-level energy efficiency optimization function is updated.

[0021] Specifically, the specific expression of the dynamic distribution function of waste heat resources is: ,in, is the ambient temperature, obtained through the PT100 sensor, , is the maximum tolerated flue gas temperature, is the weight coefficient, the high temperature section is 0.5, the medium temperature section is 0.3, and the low temperature section is 0.2.

[0022] Specifically, the gradient decision model is constructed in the following way: Establish the corresponding mapping matrix between relevant temperature and equipment ,when The i-th type of equipment is enabled when the temperature is set to zero, where i=1, 2, 3, corresponding to heat exchangers, absorption heat pumps and ORC power generation devices, and j=1, 2, 3, representing the three-level recovery stages based on the temperature threshold interval, specifically: , where the rows of the matrix represent the equipment type, the first row is the heat pipe heat exchanger, the second row is the absorption heat pump, and the third row is the ORC power generation device. The columns of the matrix represent the temperature range level, the first column is the high temperature section, the second column is the medium temperature section, and the third column is the low temperature section. The matrix logic is as follows: ,in It is the minimum startup flow threshold of the device.

[0023] Specifically, the expression of the multi-level energy efficiency optimization function is: ,in, , represents the inlet and outlet temperature difference of the i-th type equipment, is the energy consumption conversion coefficient of the corresponding equipment, heat pipe heat exchanger β1=0.8, absorption heat pump β2=1.2, ORC β3=1.5, , that is, the fixed maintenance cost, is the operating power of the i-th category equipment.

[0024] Specifically, the constraints of the multi-level energy efficiency optimization function include: Constraints on the flue gas dew point temperature threshold: ,in represents the temperature of the flue gas at the jth stage of recovery, Indicates safety margin, the safety margin is 5℃; Equipment maximum compliance constraint: The actual heat load of the i-th type of equipment needs to be less than the maximum allowable heat load of the corresponding equipment, where the maximum allowable equipment load for the heat exchanger is 2000m³ / h, the maximum allowable equipment load for the absorption heat pump is 1500m³ / h, and the maximum allowable equipment load for the ORC power generation device is 1000m³ / h; Economic constraints: ,in is the unit energy consumption cost of the i-th type of equipment, is the operating power of the i-th type equipment, is the operating time of the i-th type of equipment, is the maximum allowable cost, and the maximum cost is 500 yuan / hour.

[0025] An energy cascade utilization system based on multi-level embedded control, such as Figure 2-Figure 4 As shown, including: The multi-stage heat exchange device group includes a high-temperature section heat pipe heat exchanger (stainless steel heat pipe heat exchanger, model HRC-200, temperature resistance 400℃, heat exchange area 200 m²), a medium-temperature section absorption heat pump (lithium bromide absorption heat pump, model YHAQ-15, COP=1.8, driving heat source temperature ≥120℃), and a low-temperature section ORC power generation device (ORC generator set, model ORC-50, working fluid is R245fa, rated power 50 kW). The multi-stage heat exchange device group adopts a series-parallel hybrid arrangement. The high-temperature section and the medium-temperature section are connected in series, and a parallel bypass is set in the low-temperature section. An adjustable baffle is provided between each section. The adjustable baffle is a pneumatically adjustable baffle (model BELIMO LR24), and the flue gas flow direction is switched according to instructions; The specific design is as follows: ; Dynamic monitoring module, equipped with infrared temperature measurement array and ultrasonic flow meter, collects flue gas parameters in real time; Intelligent control unit, integrated control algorithm for intelligent control of the system, using industrial PLC, model Siemens S7-1500, CPU 1518-4 PN / DP as the main controller, real-time database storage cycle 1 second, capacity 1TB solid state hard disk, communication protocol using Modbus TCP / IP and OPC UA dual redundancy, also includes: The decision engine executes the gradient decision model to output equipment start and stop instructions, analyzes the matrix M(T,Q) every 5 seconds, and generates equipment control instructions; Optimization calculator, which solves the extreme value of multi-level energy efficiency optimization function in real time, solves the multi-level energy efficiency optimization function once every minute, and outputs the power setting value; The prediction corrector uses Kalman filtering to fuse the output of the neural network with the measured values; Model update module, which stores historical operation data and regularly optimizes neural network model parameters.

[0026] The visual human-machine interface dynamically displays the three-level recovery efficiency comparison radar chart, the multi-level energy efficiency optimization function convergence curve, and the equipment operation status topology chart. The temperature-flow trend curve supports 1-second refresh, and the equipment operation status indicator light is green-operation, yellow-standby, and red-fault.

[0027] The specific implementation is as follows: The infrared temperature measurement array evenly arranges 3 groups of sensors (2 meters apart) along the flue axis to eliminate local temperature deviations. The sampling frequency is 1Hz (1 time per second) with an accuracy of ±1°C. The ultrasonic flowmeter is installed in the straight section of the flue (10D in front and 5D in the back, D is the pipe diameter) to measure the volume flow rate with a range of 0-3000 m³ / h and an error of ±0.5%. The flue gas composition analyzer detects the H2O, O2, and CO2 content in real time and calculates the dew point temperature. The sensor data is uploaded to the PLC controller via the Modbus TCP / IP protocol. The real-time database stores historical data and supports 1-second timestamp recording. According to the temperature resistance limit of the equipment, high, medium and low temperature sections are divided, namely [200℃, 150℃], [150℃, 100℃], [100℃, 50℃]. If the predicted temperature exceeds the current range for 5 minutes, the threshold is adjusted in steps of 10℃. For example, the high temperature section is adjusted from [200℃, 150℃] to [190℃, 140℃]. The equipment start and stop priority is set. The high temperature section equipment is enabled first (heat pipe heat exchanger), and the low temperature section equipment is used as a supplement (ORC power generation device). The PLC outputs the equipment start and stop instructions according to the matrix M(T, Q). For example, when T=160℃ and Q=1200 m³ / h, the matrix is: , which means that only the high temperature section heat pipe heat exchanger is enabled; At this time, according to the equipment status started by the current matrix data, the autonomous intelligent solution to maximize the comprehensive energy efficiency is carried out, and the sequence quadratic programming algorithm is used to find the extreme value under the corresponding constraints, and the optimal operating power P of the equipment is output. The inverter frequency is adjusted according to the output power value to control the speed of the heat pump compressor or the flow rate of the ORC working fluid pump; During system operation, the efficiency deviation is calculated every 30 minutes, new data is added to the training set every week, and the model is updated to ensure that the temperature prediction error is ≤3℃ and the flow prediction error is ≤5%.

[0028] For example, in a scenario where the flue gas temperature drops suddenly and the flow rate fluctuates greatly, the initial state temperature T=170℃, the flow rate Q=1500 m³ / h, and the heat pipe heat exchanger (high temperature section) are running. At this time, the system is running and relies on the neural network model to predict that the temperature will drop to 140℃ and the flow rate will drop to 900 m³ / h in the next 15 minutes. The PLC controller obtains the corresponding instructions, starts the absorption heat pump (medium temperature section) for preheating 3 minutes in advance, turns off the heat pipe heat exchanger, and reduces the heat pump power from 25kW to 20kW to meet the economic constraints. At the same time, the actual efficiency η=73% and the theoretical value η=75% are calculated, and the weight coefficient of the medium temperature section is updated from 0.3→0.31.

[0029] The above embodiments are provided for persons familiar with the art to implement or use the present invention. Personnel familiar with the art can make various modifications or changes to the above embodiments without departing from the concept of the present invention. Therefore, the protection scope of the present invention is not limited to the above embodiments, but should be the maximum scope of the innovative features mentioned in the claims.

Claims

1. A method for cascaded energy utilization based on multi-level embedded control, characterized in that: The following steps are involved: Step S1: Real-time monitoring of the temperature T, flow rate Q and composition data of flue gas emitted by production equipment, and establishment of a dynamic distribution function f (T, Q, t) of the flue gas energy resources emitted, where t represents the time variable in the flue gas emission process; Step S2: dividing the three-stage recovery phase based on the temperature threshold interval, and determining the start and stop combination of the heat exchange equipment in each phase through the gradient decision model; Step S3: construct a multi-level energy efficiency optimization function and use a neural network to predict the change trend of flue gas energy parameters in the future period, dynamically adjust the classification threshold interval and the operating parameters of the corresponding heat exchange equipment; Step S4: The feedback adjustment module is used to compensate for the deviation between the actual recovery efficiency and the theoretical value, and the weight coefficient of the multi-level energy efficiency optimization function is updated.

2. A method for cascaded energy utilization based on multi-level embedded control according to claim 1, characterized in that: The gradient decision model is constructed in the following way: Establish the corresponding mapping matrix between relevant temperature and equipment ,when The i-th type of equipment is enabled when , where i=1, 2, 3, corresponding to the heat exchanger, absorption heat pump and ORC power generation device, and j=1, 2, 3, representing the three-level recovery stage divided based on the temperature threshold range.

3. A method for cascaded energy utilization based on multi-level embedded control according to claim 1, characterized in that: The constraints of the multi-level energy efficiency optimization function include: Constraints on the flue gas dew point temperature threshold: ,in represents the temperature of the flue gas at the jth stage of recovery, Indicates safety margin; Equipment maximum compliance constraint: The actual heat load of the i-th type of equipment needs to be less than the maximum allowable heat load of the corresponding equipment; Economic constraints: ,in is the unit energy consumption cost of the i-th type of equipment, is the operating time of the i-th type of equipment, is the maximum allowable cost, is the operating power of the i-th category equipment.

4. The energy cascade utilization method based on multi-level embedded control according to claim 1 is characterized in that: The input parameters for the neural network training include: historical operation data, ambient temperature and humidity, production equipment load rate, fuel calorific value, and the output parameters are the temperature forecast value and confidence interval for the next 15 minutes.

5. An energy cascade utilization system based on multi-level embedded control, characterized in that: Included are: A multi-stage heat exchange device group, including a high-temperature section heat pipe heat exchanger, a medium-temperature section absorption heat pump, and a low-temperature section ORC power generation device; The dynamic monitoring module is equipped with an infrared temperature measurement array and an ultrasonic flowmeter to collect the temperature, flow and composition data of the flue gas discharged by the production equipment in real time, and construct a dynamic distribution function of the flue gas energy resources based on the data. The dynamic monitoring module also includes a flue gas composition analyzer to detect the O2, CO2, and H2O content in the flue gas, calculate the flue gas dew point temperature, and combine the ambient temperature, production equipment load rate and fuel calorific value data to provide multi-dimensional input for subsequent control decisions; Intelligent control unit, integrated control algorithm to intelligently control the system; The model update module stores historical operation data and regularly optimizes the neural network model parameters to ensure prediction accuracy. The model update module calculates the efficiency deviation every 30 minutes, adds new data to the training set every week, and updates the model.

6. An energy cascade utilization system based on multi-level embedded control according to claim 5, characterized in that: The multi-stage heat exchange device group adopts a series-parallel mixed arrangement mode, the high temperature section and the medium temperature section are connected in series, the low temperature section is provided with a parallel bypass, and an adjustable baffle is provided between each section.

7. An energy cascade utilization system based on multi-level embedded control according to claim 5, characterized in that: The intelligent control unit comprises: The decision engine divides the recovery stage into three stages based on the temperature threshold interval, generates a temperature-equipment mapping matrix through a gradient decision model, and dynamically determines the start and stop combination of the heat exchange equipment in each stage. The gradient decision model generates equipment start and stop instructions based on the flue gas temperature and flow rate, combined with the equipment's minimum start flow threshold, to ensure that the equipment operates under optimal conditions; The optimization calculator constructs a multi-level energy efficiency optimization function, combines the flue gas dew point temperature constraint, the equipment maximum load constraint and the economic cost constraint, and solves the function extreme value in real time to adjust the equipment operating parameters; The prediction corrector uses a neural network model to input historical operating data, ambient temperature and humidity, and production equipment load rate to predict the changing trend of flue gas energy parameters in the future period, and performs dynamic correction by fusing the measured values ​​through Kalman filtering.

8. An energy cascade utilization system based on multi-level embedded control according to claim 5, characterized in that: It also includes a visual human-machine interface that dynamically displays a three-level recovery efficiency comparison radar chart, a multi-level energy efficiency optimization function convergence curve, and a topological diagram of the equipment operating status.

Images