CN-121980911-A - Energy pile group optimization method based on digital twin and model predictive control

CN121980911ACN 121980911 ACN121980911 ACN 121980911ACN-121980911-A

Abstract

The invention discloses an energy pile group optimization method based on digital twin and model predictive control, which predicts through a digital twin model of an energy pile group, a controller can take preventive measures before adverse effects such as thermal accumulation and the like actually occur, and changes passive response into active guidance, and the digital twin model can continuously learn and approximate the dynamic characteristics of the real energy pile group through an online correction algorithm of the digital twin model so as to improve the prediction precision and the control effect.

Inventors

WU YUANHAO
WU JIANQIU
LI ZHUOMIN
DAI XU
LI BINGXUAN
YU HAIBIN

Assignees

中国建筑第八工程局有限公司

Dates

Publication Date: 20260505
Application Date: 20251216

Claims (9)

1. An energy pile group optimization method based on digital twin and model predictive control is characterized in that, S1, constructing and updating a high-fidelity digital twin model of the energy pile group, constructing the digital twin model of the energy pile group in an analog environment based on engineering information of the energy pile group, endowing analog parameters corresponding to a physical environment, comparing real-time measured temperature field data of the energy pile group with predicted values of the digital twin model, dynamically correcting state variables of the digital twin model, S2, setting a controller based on MPC, defining a prediction time domain and a control time domain of the controller, defining an objective function to be minimized in the prediction time domain, setting optimized constraint conditions, predicting system behaviors under different control sequences in the prediction time domain by utilizing a digital twin model, solving the control sequence of the control time domain which can minimize the objective function under the condition of meeting the constraint conditions by an optimization solver, And S3, rolling execution and feedback, wherein the first step in the control sequence is issued to the energy pile group for execution, S1 and S2 are repeated at the next sampling moment, the digital twin model is updated based on the updated energy pile group temperature field data, and the optimization solution is carried out again.
2. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein the simulation parameters comprise a geotechnical thermophysical parameter, a heat exchange tube parameter and a pipeline connection relation.
3. The energy pile group optimization method based on digital twin and model predictive control according to claim 2, wherein the geotechnical thermophysical parameters include thermal conductivity, specific heat capacity and density.
4. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein temperature sensors are distributed at intervals along the depth direction of each energy pile and used for monitoring the vertical temperature gradient of the energy pile, and soil mass temperature sensing arrays are arranged in soil mass of the energy pile group area and used for monitoring the soil mass temperature so as to measure the temperature field data of the energy pile group in real time.
5. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein the energy pile group temperature field data is compared with the predicted value of the digital twin model based on a kalman filtering or particle filtering algorithm, and the state variables in the model are dynamically corrected, or the basic state variables are inverted and corrected to correct the simulation parameters.
6. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein the predictive time domain is a future time range predicted forward by the controller, the control time domain is a time range in which the controller needs to calculate a future control amount, and the control time domain is smaller than the predictive time domain.
7. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein the constraint conditions include load demand constraints, equipment capacity constraints and temperature safety constraints.
8. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein in each control period, the system behavior under different control sequences in the prediction time domain is predicted by using the digital twin model with the current system state as an initial condition.
9. The energy pile group optimization method based on digital twin and model predictive control according to claim 1, wherein the energy pile group is divided into a plurality of control areas, and an area adjusting valve is arranged on a water inlet or return main pipe of the energy pile in each control area, and is used for executing a first step in a control sequence and synchronously adjusting the flow of all the energy piles in the corresponding control area.

Description

Energy pile group optimization method based on digital twin and model predictive control Technical Field The invention relates to the technical field of underground engineering, in particular to an energy pile group optimization method. Background The support pile foundation around the subway tunnel provides a huge buried pipe space for geothermal energy development, and the energy pile group is to embed heat exchange pipes in the pile foundation, so that the heat exchange pipes exchange heat with surrounding rock-soil bodies to realize heating and refrigerating functions. However, when these dense clusters of energy piles are operated cooperatively, a strong concentrated source of thermal disturbance is formed, resulting in a sustained deviation of formation temperature in the region of the clusters of energy piles from an initial equilibrium state (e.g., sustained warm in summer and cool in winter), which is a "hot pile" effect. In the prior art, hysteresis exists in feedback control of the energy pile group, when the controller detects temperature change, heat accumulation is spread in the soil body for a certain distance, the control effect is delayed, and the effect is poor. Therefore, how to realize the operation prediction and the advanced regulation of the energy pile group, and improving the reliability of the energy pile group becomes a problem to be solved in the field. Disclosure of Invention Aiming at the defects of the prior art, the invention aims to provide an energy pile group optimization method based on digital twin and model prediction control, which can realize advanced prediction and regulation. In order to achieve the above object, the present invention provides an energy pile group optimization method based on digital twin and model predictive control, comprising: s1, constructing and updating a high-fidelity digital twin model of the energy pile group, constructing the digital twin model of the energy pile group in an analog environment based on engineering information of the energy pile group, endowing analog parameters corresponding to a physical environment, comparing real-time measured temperature field data of the energy pile group with predicted values of the digital twin model, dynamically correcting state variables of the digital twin model, S2, setting a controller based on MPC, defining a prediction time domain and a control time domain of the controller, defining an objective function to be minimized in the prediction time domain, setting optimized constraint conditions, predicting system behaviors under different control sequences in the prediction time domain by utilizing a digital twin model, solving the control sequence of the control time domain which can minimize the objective function under the condition of meeting the constraint conditions by an optimization solver, And S3, rolling execution and feedback, wherein the first step in the control sequence is issued to the energy pile group for execution, S1 and S2 are repeated at the next sampling moment, the digital twin model is updated based on the updated energy pile group temperature field data, and the optimization solution is carried out again. Further, the simulation parameters comprise rock-soil thermophysical parameters, heat exchange tube parameters and pipeline connection relations. Further, the geotechnical thermophysical parameters include a thermal conductivity coefficient, a specific heat capacity and a density. Further, temperature sensors are distributed at intervals along the depth direction of each energy pile and used for monitoring the vertical temperature gradient of the energy pile, and soil temperature sensing arrays are distributed in the soil of the energy pile group area and used for monitoring the soil temperature so as to measure the temperature field data of the energy pile group in real time. Further, based on Kalman filtering or particle filtering algorithm, comparing the temperature field data of the energy pile group with the predicted value of the digital twin model, and dynamically correcting the state variable in the model or inverting and correcting the simulation parameter based on the state variable. Further, the prediction horizon is a future time range predicted forward by the controller, the control horizon is a time range in which the controller needs to calculate a future control amount, and the control horizon is smaller than the prediction horizon. Further, constraints include load demand constraints, equipment capacity constraints, and temperature safety constraints. Further, in each control period, the current system state is used as an initial condition, and the digital twin model is used for predicting the system behavior under different control sequences in a prediction time domain. Further, the energy pile group is divided into a plurality of control areas, and an area adjusting valve is arranged on a water inlet or return main pipe of the energy pile in each contr