A predictive control algorithm for open channel water diversion projects considering large regulation interval requirements

Abstract

The present invention discloses a predictive control algorithm for a canal water diversion project considering large regulation interval requirements, which includes selecting a smaller control algorithm control interval to construct an MPC control model, and setting the weights of flow change and water level deviation; when performing real-time regulation of the canal water level, using the control interval constraint method to impose a flow amplitude change constraint when the actual desired control interval is not reached, restricting the regulation of the sluice gate, so as to achieve a longer regulation interval for the sluice gate regulation building; using the water level constraint method, setting the control weight of a smaller water level deviation interval to 0; the purpose of the present invention is to provide a method for increasing the control interval, extending the gate regulation interval time, and reducing the number of gate regulations by using the control interval constraint and water level constraint methods without reducing the control effect of the predictive control algorithm, and meeting the requirements of water level safety regulation in large canal water diversion projects and minimizing the participation of sluice gates in regulation through this algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time regulation of open channel water transfer projects, and in particular to a predictive control algorithm for open channel water transfer projects that takes into account large regulation interval requirements. Background Art

[0002] With the development of social economy, people's demand for water resources is increasing. The construction of large-scale long-distance water transfer projects has become the most effective and direct means to regulate the uneven spatial and temporal distribution of water resources and solve the contradiction between water supply and demand. At present, most of the large-scale open channel water transfer projects built are open channel water transfer projects for the purpose of ensuring water supply. That is, the main purpose of water transfer is to meet the water needs of water users at various water diversion outlets downstream. In order to control the change of channel water level, it is necessary to solve the gate control problem of the water supply channel to achieve stable water level and stable water supply, while the water level does not exceed the allowable range to avoid causing serious harm.

[0003] For open channels, the water level control of the channel is mainly completed by the regulating gate. For the channel pool, which aims to ensure water supply, the upstream gate is generally adjusted to keep the downstream flow unchanged as much as possible and meet the downstream water supply. Therefore, the current open channel water transfer project is moving towards the direction of unmanned and automated operation. The design and research of the automated control system of the channel pool has also made great progress.

[0004] At present, several technologies have been developed to handle the gate control problem of canal pools. The current canal control methods can be roughly divided into feedback method, feedforward method and model predictive control (MPC) method. The basic idea of the feedback control method is to obtain the adjustment rules or logics of the control structure. The most representative one is the canal pool control system that directly uses the classical proportional-integral (PI) controller for gate control. PI control is the most widely used control method in practice because the rules or logics are easy to calculate, but it is usually only used to guide a limited number of gates in the canal pool. When many gates in a multi-stage series canal pool are controlled, it is difficult to form the control logic; the most famous one in the feedforward method is the capacity compensation method. However, due to the influence of the uncertainty of pipeline parameters, the capacity compensation needs to be applied wisely and is preferably used together with the feedback control strategy; for the model predictive control algorithm, compared with other algorithms, it can better predict the canal pool state and thus adjust the optimal control strategy in real time, and has better canal pool water level control effect. The model predictive control algorithm requires an internal model to predict the state of the research object. The current internal model mainly uses the integrator time delay model to construct the internal model. However, due to some characteristics of the integrator time delay model, a small control interval is required to maintain the prediction accuracy, and a large regulation interval cannot be used for the discretization of the internal model, resulting in a relatively small control interval for the model predictive control algorithm constructed based on this. Under the action of the model predictive control algorithm, it will lead to frequent gate control. In some large-scale water diversion channels, if the gate control is too frequent, it may damage the large control devices. In addition, in some large-scale water diversion projects with pump stations as the main control devices, due to the small adjustability of the pump stations, the regulation should not be too frequent. Therefore, it is necessary to develop a real-time control method that not only meets the water level stability requirements but also limits the gate control frequency. Summary of the Invention

[0005] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract and the title, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0006] In view of the above problems existing in the model predictive control algorithm for open channel water diversion projects considering the large regulation interval requirements, the present invention is proposed.

[0007] Therefore, the purpose of the present invention is to provide a model predictive control algorithm for open channel water diversion projects considering the large regulation interval requirements, which meets the water level stability requirements and limits the gate control frequency.

[0008] To solve the above technical problems, the present invention provides the following technical solutions: A predictive control algorithm for a canal water diversion project considering large regulation interval requirements, including constructing an initial predictive control model, establishing an integral time-delay model and its constraints; establishing a control interval constraint method, calculating the multiple of the actual required regulation interval and the control interval of the control system and setting constraints; constructing a water level constraint method.

[0009] As a preferred embodiment of the predictive control algorithm for a canal water diversion project considering large regulation interval requirements according to the present invention, wherein: The gate control model assumes that the water level deviation at the downstream end is e, which is a linear function of the flow rate change, and the upstream flow rate change has a delay effect on the water level at the downstream end of the canal pond; The calculation of the integral time-delay (ID) model can be expressed as:

[0010]

[0011] wherein, t d represents the time delay; q in (t - t d ) is the deviation of the upstream inflow of the backwater section with a lag time of t d from the initial flow rate; q out (t) is the deviation of the downstream outflow from the initial flow rate; q offtake (t) is the deviation of the water diversion flow rate from the initial flow rate; A s is the water surface area.

[0012] As a preferred embodiment of the predictive control algorithm for a canal water diversion project considering large regulation interval requirements according to the present invention, wherein: The integral time-delay model is written in a discrete form:

[0013]

[0014]

[0015] wherein, Δe(k) = e(k) - e(k - 1) is the increment of e(k) relative to the previous step; Similarly, Δq in (k), Δq out (k) and Δq offtake (k) are the increments of q in (k), q out (k) and q offtake (k) relative to the previous step; When writing the ID model into a discrete expression, the delay time will be converted into a delay step, and the formula is as follows:

[0016] k d = t d / t s (4)

[0017] Among them, the delay step k d is an integer, and t s is the control interval in the control system; in order to maintain an accurate description of the model time-delay characteristics, the value of the control interval t s should be less than or equal to t d .

[0018] As a preferred solution of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirements of the present invention, wherein: Formulas (2) and (3) are written in the following form:

[0019] x(k + 1) = Ax(k) + Bu(k) + Dd(k) (5)

[0020] y(k) = Cx(k) (6)

[0021] Among them, x(k) is the state variable; y(k) is the output variable; u(k) is the control input variable; d(k) is the disturbance variable; the control A is the system matrix; B is the control matrix; D is the disturbance matrix; C is the output matrix in the form of a diagonal matrix.

[0022] As a preferred solution of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirements of the present invention, wherein: Assume that the water level deviation at the downstream point of the canal pond i is e i , and e i,r is the reference trajectory of e i . To keep the water level stable, e i,r can be set to zero, and ΔQ i is the increment of the upstream inflow of the canal pond i; then the state space model of the ID model is used for water level deviation prediction; the objective function of MPC in water level control is expressed as:

[0023]

[0024] In predictive control, the quadratic programming algorithm is generally used for solution, and its constraint form needs to satisfy:

[0025] u lb (k + j - 1|k) ≤ u(k + j - 1|k) ≤ u ub (k + j - 1|k) (8).

[0026] As a preferred solution of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirements of the present invention, wherein: Determine the required regulation interval according to whether the required regulation interval is L times the regulation interval used by the internal model:

[0027] T req / t s ≤ L < T req / t s+1 (9)

[0028] where L is an integer.

[0029] As a preferred solution of the predictive control algorithm for the open-channel water transfer project considering the large regulation interval requirement of the present invention, wherein: if the control step is not a multiple of L, the flow control action ΔQ(k) in the control variable u(k) can be strictly limited to 0; this flow change constraint can be added to the constraint equation:

[0030] 0≤ΔQ(k+j-1|k)≤0 if(k+j-1)%L≠0 (10)

[0031] The constraint of the control variable is still in the form of constraint (8), so the solution method can still be quadratic programming, and the flow control interval of this constraint can be increased to the regulation interval T required by the actual project req .

[0032] As a preferred solution of the predictive control algorithm for the open-channel water transfer project considering the large regulation interval requirement of the present invention, wherein: the water level constraint is usually used to limit the controlled water level within the allowable range, by introducing a virtual input signal u in the input vector u * and introducing a virtual state e in the state vector x * , using this constraint to limit the water level during a significant change in the water flow; assuming e max and e min are the upper and lower values of the water level deviation constraint respectively, the virtual state can be expressed by the following formula:

[0033] e * (k)=e(k)-u * (k) (11)

[0034] where:

[0035]

[0036] As a preferred solution of the predictive control algorithm for the open-channel water transfer project considering the large regulation interval requirement of the present invention, wherein: after using this constraint, the deviation part exceeding the water level deviation constraint is set as the virtual state e * (k), and if a weight is given to this e * (k), it will additionally weight and punish the deviation of the part exceeding the constraint. Therefore, the weight of the part exceeding the allowable range is the sum of the weight e(k) and the weight e * (k); by setting a significant weight for e * (k), the water level should be kept as far as possible within the specified range.

[0037] As an optimal solution of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirement of the present invention, among others: in addition to setting constraints for the allowable amplitude variation, several more intervals can also be set within the allowable range to set constraints: for e max and e min a range is further set inside, e min2 and e max2 , where e min2 is greater than e min , and e max2 is less than e max . For the water level deviation within the range between e min2 and e max2 , the weight of the water level deviation is set to zero, so that the tiny water level deviation is not controlled or is slightly controlled.

[0038] Advantages of the present invention:

[0039] The purpose of the present invention is to provide a method that, without reducing the control effect of the predictive control algorithm, uses the control interval constraint and water level constraint to increase the control interval, extend the gate regulation interval time, and reduce the number of gate regulations, so as to meet the requirements of water level safety regulation in large open-channel water diversion projects and minimize the participation of check gates in regulation through this algorithm. Description of the drawings

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0041] Figure 1 is a schematic diagram of the research channel pond of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirement of the present invention.

[0042] Figure 2 is a graph of the absorption flow rate change of the unknown minor transportation disturbance of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirement of the present invention.

[0043] Figure 3 is a graph of the gate opening control result of the MPC control method with a small control interval and no constraints for unknown disturbances of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirement of the present invention.

[0044] Figure 4 is a graph of the gate opening control result of the MPC control method with two constraints for unknown disturbances of the predictive control algorithm for the open-channel water diversion project considering the large regulation interval requirement of the present invention. Detailed implementation manners

[0045] To make the above objects, features and advantages of the present invention more obvious and understandable, the following detailed description of the specific implementation manners of the present invention will be given in conjunction with the accompanying drawings of the specification.

[0046] In the following description, many specific details are set forth to facilitate a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0047] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure or characteristic that may be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that exclude each other with other embodiments.

[0048] Thirdly, the present invention will be described in detail in conjunction with the schematic diagrams. When detailing the embodiments of the present invention, for the sake of convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally in a non-general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width and depth should be included.

[0049] Embodiment 1

[0050] Referring to Figures 1 to 4 , as the first embodiment of the present invention, a prediction control algorithm for a canal water diversion project considering large regulation interval requirements is provided, taking 13 series-connected water conveyance channels of a certain project as an example. The lengths of the respective canal pools in the research area vary from 10 kilometers to 27 kilometers, and there is a target water level upstream of each check gate. During the regulation process of the project, it is generally required that the regulation intervals of each inspection gate be greater than 2 hours as much as possible. Under normal operating conditions, the water level deviation of canal pools 1 to 12 should be kept less than 0.15 m, and the water level deviation of canal pool 13 should be less than 0.1 m. In case of an emergency, the water levels of canal pools 1 to 12 should be kept less than 0.3 m, and the water level deviation of canal pool 13 should be less than 0.15 m.

[0051] Step 1: 1) Construct an MPC model

[0052] The simulation models of each channel are established separately, and the measured values of the downstream water levels of the check gates are used as the model boundaries. Using the simulation model of each pool, through the model simulation results, the parameters of each integral time-delay model are obtained statistically for the canal pools. The initial flow conditions and model parameters of each pool are shown in Table 1. These pools were initially stable, and the water levels at the control points were all at the target water levels.

[0053] Table 1. Initial conditions of each pool and ID model parameters

[0054]

[0055] Then, the control method was implemented on this simulation model. The control method was tested using an unknown slight conveying disturbance method. For the unknown slight disturbance, the flow rate change information in the ID model is as Figure 2 shown.

[0056] When using the MPC control method with a small control interval and no constraints, the weight of the flow rate change in Rj was set to 1, and the weight of the water level deviation was set to 13.

[0057] Step 2: 1) Use the MPC control method with control interval constraints

[0058] When using the MPC control method with control interval constraints, the delay time range of the ID model was from 19 min to 76 min. Considering the accuracy of the internal model and the computational workload of the control method, the control interval was set to 0.5 h. Since the desired interval was 2 h, the value of L was set to 4 under the control interval constraint.

[0059] Step 3: 1) Adopt the MPC control method with water level constraints

[0060] When using the MPC control method with the water level constraint method, three virtual input signals and the corresponding virtual deviation states e* 1 , e* 2 , e* 3 were used. When the absolute value of the water level deviation of pools 1 to 12 was greater than 0.08 m or the absolute value of the water level deviation of pool 13 was greater than 0.05 m, there was a virtual deviation state e* 1 in each pool; when the absolute value in pools 1 to 12 was greater than 0.15 m or the absolute value in pool 13 was greater than 0.1 m, there was a virtual deviation state e* 2 in each pool; when the absolute value of pools 1 to 12 was greater than 0.35 m or the absolute value of pool 13 was greater than 0.15 m (this was regarded as an emergency), there was a virtual deviation state e* 3 in each pool. The weight of the water level deviation e was set to 0, while the weights of e* 1 , e* 2 and e* 3 were set to 25, 25, and 25 respectively; therefore, minor water level deviations were not controlled.

[0061] In this paper, the calculation results of the unconstrained MPC control method (Method 1) and the MPC control method using two constraints simultaneously (Method 2) are shown in Table 2; the calculation results of the water level deviation and gate opening of the unconstrained MPC control method are shown inFigure 2 Using the MPC control method with control interval constraints and water level constraints, the calculation results of water level deviation and gate opening are shown respectively in Figure 3 and Figure 4 .

[0062] Table 2. Number of gate controls for different regulation methods

[0063]

[0064]

[0065] It can be seen from Figure 3 and Figure 4 that when the control intervals of several surrounding pools are small and the deviations are large, the gate control strategy will be executed in each control interval, resulting in large fluctuations in the water level. These fluctuations will interfere with the prediction of the water level change trend, leading to unreasonable opening and closing adjustments before and after the gate. When the two constraint methods are adopted, the control interval of each gate is larger than that of the traditional MPC control method. The number of gate regulations is reduced from 139 times to 69 times, a reduction of 60%. These constraint methods achieve the purpose of increasing the gate control interval and reducing the number of gate regulations in the canal pool under unknown minor disturbances, and avoiding unnecessary control of small water level deviations.

[0066] 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 preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A predictive control algorithm for open channel water transfer projects considering large regulation interval requirements, characterized in that: It includes, Select a smaller control algorithm control interval to construct an MPC control model, and set the flow change weight and water level deviation weight; When performing real-time regulation of the channel water level, use the control interval constraint method to impose a flow amplitude change constraint when the actual desired control interval is not reached, restricting the regulation of the sluice gate, so as to achieve a longer regulation interval for the sluice gate regulation building; Use the water level constraint method to set the control weight of a smaller water level deviation interval to 0; Assume that the water level deviation at the downstream point of canal pond i is e i , e i,r is the reference trajectory of e i . To maintain the water level stability, e i,r can be set to zero, and ΔQ i is the increment of the upstream inflow of canal pond i. Then, the state - space model of the ID model is used for water level deviation prediction. The objective function of MPC in water level control is expressed as: In predictive control, use the quadratic programming algorithm to solve, and its constraint form needs to satisfy: u lb (k + j - 1|k) ≤ u(k + j - 1|k) ≤ u ub (k + j - 1|k) (8); Determine the required regulation interval according to whether the required regulation interval is L times the regulation interval used by the internal model: T req / t s ≤L<T req / t s +1 (9) where L is an integer; If the control step is not a multiple of L, the flow control action ΔQ(k) in the control variable u(k) can be strictly restricted to 0; this flow change constraint can be added to the constraint equation: 0≤ΔQ(k+j-1|k)≤0 if (k+j-1)%L≠0 (10) The constraint of the control variable remains in the form of constraint (8), so that the solution method can still be quadratic programming, and the flow control interval of this constraint can be increased to the regulation interval T required by the actual project req ; Water level constraints are usually used to limit the controlled water level within an allowable range by introducing a virtual input signal u in the input vector u * and a virtual state e in the state vector x * , and this constraint is used to limit the water level during significant changes in water flow; assuming that e max and e min are the upper and lower values of the water level deviation constraint respectively, the virtual state can be expressed by the following formula: e * e(k) = e(k) - u * e(k) (11) where: After using this constraint, the deviation part exceeding the water level deviation constraint is set to the virtual state e * (k), and if weights are also assigned to this e * (k), additional weighted penalties will be imposed on the deviation of the part exceeding the constraint. Therefore, the weight of the part exceeding the allowable range is the sum of the weight e(k) and the weight e * (k); by setting a significant weight for e * (k), the water level should be kept as far as possible within the specified range.

2. The predictive control algorithm for open channel water transfer projects considering large regulation interval requirements according to claim 1, characterized in that: The gate control model assumes that the water level deviation at the downstream end is e, which is a linear function of the flow change, and the upstream flow change has a delay effect on the water level at the downstream end of the canal pond; the calculation of the integral time delay (ID) model can be expressed as: where t d represents the time delay; q in (t - t d ) is the deviation of the flow rate flowing into the upstream of the backwater section with a lag time of t d from the initial flow rate; q out (t) is the deviation of the flow rate flowing out downstream from the initial flow rate; q offtake (t) is the deviation of the diversion flow rate from the initial flow rate; A s is the water surface area.

3. The predictive control algorithm for open channel water transfer projects considering large regulation interval requirements according to claim 2, characterized in that: Write the integral time delay model in discrete form: where, Δe(k) = e(k) - e(k - 1) is the increment of e(k) relative to the previous step; similarly, Δq in (k), Δq out (k) and Δq offtake (k) are the increments of q in (k), q out (k) and q offtake (k) relative to the previous step respectively; when writing the ID model into the discrete expression, the delay time will be converted into the delay step, and the formula is as follows: k d = t d / t s (4) Among them, the delay step k d is an integer, and t s is the control interval in the control system; in order to maintain an accurate description of the model time-delay characteristics, the value of the control interval t s should be less than or equal to t d .

4. The predictive control algorithm for open channel water transfer projects considering large regulation interval requirements according to claim 3, characterized in that: Write formulas (2) and (3) in the following form: x(k+1)=Ax(k)+Bu(k)+Dd(k) (5) y(k)=Cx(k) (6) where x(k) is the state variable; y(k) is the output variable; u(k) is the control input variable; d(k) is the disturbance variable; control A is the system matrix; B is the control matrix; D is the disturbance matrix; C is the output matrix in diagonal matrix form.

5. The predictive control algorithm for open channel water transfer projects considering large regulation interval requirements according to claim 4, characterized in that: In addition to setting constraints for the allowable amplitude variation, several more intervals can be set within the allowable range to set constraints: at e max and e min an additional range is set inside, e min2 and e max2 , where e min2 is greater than e min , and e max2 is less than e max . For the water level deviation within the range between e min2 and e max2 , the weight of the water level deviation is set to zero, so that the minute water level deviation is not controlled or is slightly controlled.

Images