An adaptive path planning method for urban sustainable development target

Abstract

This invention discloses an adaptive path planning method for urban sustainable development goals, belonging to the field of resource planning and management technology. This method maps secondary indicators with different directions and dimensions to a unified score domain and calculates the overall index. It uses a four-endpoint product propagation method relying only on the four arithmetic operations of interval endpoints to obtain the terminal reachable intervals year by year, thereby determining the effective target index. Under the constraint that the weighted sum of the primary target scores equals the effective target index, it solves for the equilibrium anchor point, obtaining each emphasis endpoint through a two-stage lexicographical solution. The endpoints are then inverted into secondary indicator terminal values, and feasible values ​​are selected year by year from the intersection of forward and backward reachable intervals to form multi-scenario paths. Each path is independently verified, and failed paths are automatically downgraded. This method avoids solution failures caused by unreachable targets, improving the feasibility, separation, and computational efficiency of multi-scenario paths.

Description

Technical Field

[0001] This invention belongs to the field of resource planning and management technology, and specifically relates to an adaptive path planning method oriented towards urban sustainable development goals. Background Technology

[0002] The evaluation of Sustainable Development Goals typically consists of multiple primary goals and several secondary indicators belonging to each primary goal. Different secondary indicators have different dimensions such as percentage, population, concentration, amount, and area, and the direction of improvement may be positive or negative. Regional planning is also subject to the combined constraints of policy thresholds, natural conditions, historical rate of change, and statistical methods, thus requiring a reliable mapping between a unified score space and the original indicator space.

[0003] Existing methods often focus on statically scoring historical states or directly invoking a general nonlinear optimizer to generate a single path given a specific objective. When the target year is recent, the target index is high, or some indicators are close to the boundary, direct optimization is prone to entering the infeasible region, resulting in solution failure, anomalous jumps in the first planning year, out-of-bounds terminal values, or conflicts in annual rate of change. If decision variables are established simultaneously for each planning year and each secondary indicator, the decision dimensionality will increase multiplied by the number of indicators and the planning period, leading to a significant increase in time consumption.

[0004] On the other hand, planning users not only need a balanced development path, but also need to compare multiple emphase scenarios. Simply requiring the score of the emphased first-level objective to be higher than the scores of all other first-level objectives will introduce unnecessary hard conflicts in some regions or indicator systems; if the emphase constraint is completely removed, the endpoints of each scenario are likely to converge to the same equilibrium solution, resulting in insufficient path separation. Therefore, a multi-scenario method is needed that can first identify the reachable extrema of the emphased objective, and then construct a stable endpoint within the neighborhood of the extrema.

[0005] Furthermore, the number of indicators, the number of primary objectives, the aggregation method, and the number of objectives without data can all change dynamically, making it difficult to scale dedicated algorithms with fixed dimensions or manual configuration. Therefore, there is an urgent need for an implementation scheme that is decoupled from the scale of the indicator system, can automatically determine the achievability of objectives, and can verify them path by path. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes an adaptive path planning method oriented towards urban sustainable development goals. This method can solve the problems of existing technologies, such as easy solution failure when the goal is unreachable with a finite number of steps, insufficient path separation in multiple scenarios, large computational load in full-time nonlinear optimization, and lack of independent verification of path constraints.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: An adaptive path planning method oriented towards urban sustainable development goals includes the following steps: S1. Obtain the hierarchical indicator system configuration, current status year, requested target index, target year, and parameters of each secondary indicator for the area to be planned; the parameters of the secondary indicators include the current value, best value, worst value, legal lower bound, legal upper bound, minimum annual change rate, and maximum annual change rate; S2. Based on the best and worst values, map the secondary indicators with different directions and dimensions to a unified score domain, and obtain the scores of each primary target and the current overall index according to the configured aggregation method. S3. Starting from the current value of the secondary indicator, propagate the reachable range of each secondary indicator year by year based on the annual minimum change rate, annual maximum change rate, legal lower bound, and legal upper bound to obtain the terminal reachable range corresponding to the target year. S4. Obtain the highest reachable value of the overall index based on the terminal reachable range. Output an input error when the requested target index is lower than the current overall index. Determine the highest reachable value as the valid target index when the requested target index is higher than the highest reachable value. Otherwise, determine the requested target index as the valid target index. S5. Under the constraint that the weighted sum of the scores of the first-level objectives equals the effective objective index and each score of the first-level objectives is within its respective reachable interval, the equilibrium anchor point is solved with the deviation of each score of the first-level objectives from the effective objective index as the objective function. S6. For each primary objective with effective secondary indicators, first solve for the maximum score of the primary objective under the same effective objective constraints as in step S5, then keep the score of the primary objective within the preset neighborhood of the maximum score, and solve for the focus endpoint with the goal of minimizing the disturbance of the relatively balanced anchor point. S7. Invert the equilibrium anchor point and each emphasis endpoint into the terminal value of the secondary indicator. Based on the intersection of the forward reachable interval and the backward reachable interval derived from the terminal value, select feasible secondary indicator values ​​year by year to form the equilibrium path and multiple emphasis paths. S8. Verify the numerical finiteness, legal upper and lower bounds, annual change rate, and overall index residual of the endpoint for each path, and output the first-level target path, second-level indicator path, overall index path, and coupling coordination degree path that have passed the verification. S9. If a path fails the verification, record the primary target number and the year of violation corresponding to the path, and automatically downgrade to use the feasible path from the previous iteration as the alternative output.

[0008] Preferably, the specific process of step S2 is as follows: S21. Map the secondary indicators to a unified score range. The calculation formula is as follows: ,in, is the standardized score of the raw value of the j-th secondary indicator; j is the index of the secondary indicator, j=1,2,…,N, and N is the total number of valid secondary indicators; This represents the original value of the j-th secondary indicator; This represents the best value of the j-th secondary indicator; Let be the worst value of the j-th secondary indicator, and Not equal to ; Π [0,100] For projection operators onto the closed interval [0, 100], results less than 0 are taken as 0, and results greater than 100 are taken as 100; for positive indices, > For negative indicators, < ; S22. The formula for calculating the score of the first-level objective is: ,in, The score is the score of the g-th primary objective; g is the primary objective number, g=1,2,…,K, and K is the total number of primary objectives with at least one valid secondary indicator; This is the set of valid secondary indicator indices belonging to the g-th primary objective; The non-negative aggregate weight of the j-th secondary indicator; primary objectives without effective secondary indicators are not included in the K, overall index, and optimization variable set, and are marked as having no data in the output; The current formula for calculating the overall index is: Where S is the current overall index; The normalized non-negative weights for the g-th first-level objective; each The sum is 1.

[0009] Preferably, in step S3, the j-th secondary index is propagated using a four-endpoint product propagation method in year t, letting... , ,in and The candidate set consists of the minimum annual rate of change and the maximum annual rate of change. ,in, Let j be the candidate set of the four endpoints of the j-th secondary indicator in year t; and These represent the lower and upper ends of the reachable range of the j-th secondary indicator in year t-1, respectively. and These represent the lower and upper ends of the annual multiplicative factor for the j-th secondary indicator, respectively; t represents the year. The formula for calculating the reachable interval in year t is: ,in, and These represent the lower and upper ends of the reachable interval of the j-th secondary indicator in year t, respectively. and Let $\mathbf{j}$ be the legal lower bound and legal upper bound of the $j$-th secondary index, respectively; $\min(·)$ and $\max(·)$ are the minimum and maximum value operations, respectively; the initial condition is... , This represents the original value of the j-th secondary indicator in the current year.

[0010] Preferably, in step S4, the number of planning steps is equal to the difference between the target year and the current state year, and the formula for calculating the effective target index is: Where T is the number of planning steps; The target year set by the user; The current state year; For effective target index; The target index for the user's request; This represents the highest achievable value of the overall index within the planning steps T. When T≤0, output "Target Year Error"; when the requested target index is lower than the current overall index, output "Target Index Lower than Current Overall Index"; where the current overall index is the overall index calculated from the current state.

[0011] Preferably, in step S5, the formula for calculating the equilibrium anchor point is: , Where B is the equilibrium anchor point; s is the first-level target terminal score vector; and These are the lower and upper bounds of the terminal reachability score for the g-th primary objective, respectively.

[0012] Preferably, in step S6, the k-th primary objective is solved using a two-stage lexicographical order, with the first stage being: The first stage uses the same first-level objective reachability score boundary constraints and effective objective exponential equality constraints as step S5; the second stage is: , Where k is the index of the primary target currently being emphasized, k∈{1,2,…,K}; This represents the highest attainable score of the k-th primary objective under the effective objective index constraint. The terminal score for the k-th primary objective; Let be the stable endpoint vector of the extreme neighborhood of the k-th primary objective; The g-th first-order target component of the equalization anchor point; ε is the non-negative extreme value neighborhood tolerance; no tolerance is set in the second stage. A mandatory dominance constraint that is higher than the scores of all other first-level objectives.

[0013] Preferably, in step S6, the formula for calculating the kth emphasis endpoint of the adjustable separation degree is: ,in, This is the primary endpoint of the k-th primary objective; This is the path separation coefficient; since B and All are located on the same effective target hyperplane. Keep the overall index equal to .

[0014] Preferably, in step S7, for each secondary indicator, the reachable intervals for each year obtained from the forward propagation of the current value are first saved, and then the backward reachable intervals for each year are obtained from the terminal value according to the annual rate of change. In the intersection of the forward reachable interval, backward reachable interval and legal upper and lower bounds of the corresponding year, the annual value is selected according to the preset smoothing progress. The secondary indicators to which the primary target belongs and the secondary indicators not to which the primary target belongs are offset by opposite progress biases to enhance the recognizability of the emphasized path on the time axis.

[0015] Preferably, in step S8, the formulas for calculating the coupling degree, the coordination comprehensive index, and the coupling coordination degree are as follows: ,in, Coupling degree; Score for the g-th primary objective The dimensionless normalized value, And 0≤ ≤1; denoted as the comprehensive coordination index; D represents the coupling coordination degree.

[0016] After adopting the above technical solution, the present invention has the following beneficial effects: First, by decomposing high-dimensional full-time domain optimization into terminal reachability interval propagation, low-dimensional first-level objective allocation, and analytical path reconstruction, the present invention can significantly reduce the number of variables processed by the general nonlinear optimizer; Second, by using four-endpoint product propagation, the present invention accurately determines the reachability of a finite number of steps before formal solution, and automatically provides the highest reachability boundary solution when it is unreachable, which helps to avoid long-term failed iterations; Third, the present invention adopts a two-stage method focusing on extreme values ​​and extreme value neighborhood stabilization, which can maintain the separation degree of multiple scenarios without introducing global dominance conflicts; Fourth, the present invention generates paths by the intersection of forward and backward reachability intervals, which helps to avoid abnormal jumps in the first planning year; Fifth, the present invention uses independent constraint verification and structured diagnosis, which can improve the reliability of results and the maintainability of the system; Sixth, the present invention supports dynamic index systems and data-free objectives, which is easy to extend to different regions. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention. 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 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 As shown. An adaptive path planning method oriented towards urban sustainable development goals includes the following steps: S1. Obtain the hierarchical indicator system configuration, current status year, requested target index, target year, and parameters of each secondary indicator for the area to be planned; the parameters of the secondary indicators include the current value, best value, worst value, legal lower bound, legal upper bound, minimum annual change rate, and maximum annual change rate; S2. Based on the best and worst values, map the secondary indicators with different directions and dimensions to a unified score domain, and obtain the scores of each primary target and the current overall index according to the configured aggregation method. The specific process of step S2 is as follows: S21. Map the secondary indicators to a unified score range. The calculation formula is as follows: ,in, is the standardized score of the raw value of the j-th secondary indicator; j is the index of the secondary indicator, j=1,2,…,N, and N is the total number of valid secondary indicators; This represents the original value of the j-th secondary indicator; This represents the best value of the j-th secondary indicator; Let be the worst value of the j-th secondary indicator, and Not equal to ; Π [0,100] For projection operators onto the closed interval [0, 100], results less than 0 are taken as 0, and results greater than 100 are taken as 100; for positive indices, > For negative indicators, < ; S22. The formula for calculating the score of the first-level objective is: ,in, The score is the score of the g-th primary objective; g is the primary objective number, g=1,2,…,K, and K is the total number of primary objectives with at least one valid secondary indicator; This is the set of valid secondary indicator indices belonging to the g-th primary objective; The non-negative aggregate weight of the j-th secondary indicator; primary objectives without effective secondary indicators are not included in the K, overall index, and optimization variable set, and are marked as having no data in the output; The current formula for calculating the overall index is: Where S is the current overall index; The normalized non-negative weights for the g-th first-level objective; each The sum is 1; S3. Starting from the current value of the secondary indicator, propagate the reachable range of each secondary indicator year by year based on the annual minimum change rate, annual maximum change rate, legal lower bound, and legal upper bound to obtain the terminal reachable range corresponding to the target year. In step S3, the j-th secondary index is propagated using a four-endpoint product propagation method in year t, let... , ,in and The candidate set consists of the minimum annual rate of change and the maximum annual rate of change. ,in, Let j be the candidate set of the four endpoints of the j-th secondary indicator in year t; and These represent the lower and upper ends of the reachable range of the j-th secondary indicator in year t-1, respectively. and These represent the lower and upper ends of the annual multiplicative factor for the j-th secondary indicator, respectively; t represents the year. The formula for calculating the reachable interval in year t is: ,in, and These represent the lower and upper ends of the reachable interval of the j-th secondary indicator in year t, respectively. and Let $\mathbf{j}$ be the legal lower bound and legal upper bound of the $j$-th secondary index, respectively; $\min(·)$ and $\max(·)$ are the minimum and maximum value operations, respectively; the initial condition is... , This represents the original value of the j-th secondary indicator in the current year. S4. Obtain the highest reachable value of the overall index based on the terminal reachable range. Output an input error when the requested target index is lower than the current overall index. Determine the highest reachable value as the valid target index when the requested target index is higher than the highest reachable value. Otherwise, determine the requested target index as the valid target index. In step S4, the number of planning steps equals the difference between the target year and the current state year. The formula for calculating the effective target index is: Where T is the number of planning steps; The target year set by the user; The current state year; For effective target index; The target index for the user's request; This represents the highest achievable value of the overall index within the planning steps T. Output an error message indicating that the target year is lower than the current overall index when T ≤ 0; output an error message indicating that the target index is lower than the current overall index when the requested target index is lower than the current overall index; where the current overall index is the overall index calculated from the current state; S5. Under the constraint that the weighted sum of the scores of the first-level objectives equals the effective objective index and each score of the first-level objectives is within its respective reachable interval, the equilibrium anchor point is solved with the deviation of each score of the first-level objectives from the effective objective index as the objective function. In step S5, the formula for calculating the equilibrium anchor point is: , Where B is the equilibrium anchor point; s is the first-level target terminal score vector; and These are the lower and upper bounds of the terminal reachability score for the g-th primary objective, respectively. S6. For each primary objective with effective secondary indicators, first solve for the maximum score of the primary objective under the same effective objective constraints as in step S5, then keep the score of the primary objective within the preset neighborhood of the maximum score, and solve for the focus endpoint with the goal of minimizing the disturbance of the relatively balanced anchor point. In step S6, the k-th primary objective is solved using a two-stage lexicographical order approach. The first stage is: The first stage uses the same first-level objective reachability score boundary constraints and effective objective exponential equality constraints as step S5; the second stage is: , Where k is the index of the primary target currently being emphasized, k∈{1,2,…,K}; This represents the highest attainable score of the k-th primary objective under the effective objective index constraint. The terminal score for the k-th primary objective; Let be the stable endpoint vector of the extreme neighborhood of the k-th primary objective; The g-th first-order target component of the equalization anchor point; ε is the non-negative extreme value neighborhood tolerance; no tolerance is set in the second stage. A mandatory dominance constraint that is higher than the scores of all other level-one objectives; In step S6, the formula for calculating the kth emphasis endpoint of the adjustable separation degree is: ,in, This is the primary endpoint of the k-th primary objective; This is the path separation coefficient; since B and All are located on the same effective target hyperplane. Keep the overall index equal to ; S7. Invert the equilibrium anchor point and each emphasis endpoint into the terminal value of the secondary indicator. Based on the intersection of the forward reachable interval and the backward reachable interval derived from the terminal value, select feasible secondary indicator values ​​year by year to form the equilibrium path and multiple emphasis paths. In step S7, for each secondary indicator, the reachable intervals for each year obtained from the forward propagation of the current value are first saved, and then the backward reachable intervals for each year are obtained from the terminal value according to the annual rate of change. In the intersection of the forward reachable interval, backward reachable interval and legal upper and lower bounds of the corresponding year, the annual value is selected according to the preset smoothing progress. The secondary indicators to which the primary target belongs and the secondary indicators not to which the primary target belongs are given opposite progress biases to enhance the recognizability of the focus path on the time axis. S8. Verify the numerical finiteness, legal upper and lower bounds, annual change rate, and overall index residual of the endpoint for each path, and output the first-level target path, second-level indicator path, overall index path, and coupling coordination degree path that have passed the verification. In step S8, the formulas for calculating the coupling degree, coordination comprehensive index, and coupling coordination degree are as follows: ,in, Coupling degree; Score for the g-th primary objective The dimensionless normalized value, And 0≤ ≤1; D represents the coordination comprehensive index; D represents the coupling coordination degree. S9. If a path fails the verification, record the primary target number and the year of violation corresponding to the path, and automatically downgrade to use the feasible path from the previous iteration as the alternative output.

[0020] Example 1: Regional Planning within the United Nations Sustainable Development Goals Indicator System The following request parameters are accepted: the indicator system type is UN SDGs, the demonstration area name is Guilin, the target index is 82.00, the target year is 2030, and coverage of the legal upper and lower bounds of individual secondary indicators is allowed. The current state year of the region configuration is 2024, therefore the planning step T is 6.

[0021] The configuration files are stored hierarchically according to primary objectives and secondary indicators. Taking primary objective 11, "Sustainable Towns," as an example, its secondary indicator 11.8 is "Percentage of built-up areas meeting sponge city construction requirements," in percentage (%). The current value is 40.69, and historical valid data include 21.13, 30.03, 37.14, and 40.69. The legal lower bound is set at 21.13, and the legal upper bound is set at 100. The system does not recognize year numbers as indicator values ​​and removes records that are significantly inconsistent with the indicator units and adjacent sequences during the data preprocessing stage.

[0022] The computing device first standardizes all valid secondary indicators. For proportional indicators where higher values ​​are better, the best value can be set to 100, based on UN target guidance, national standards, or policy objectives, while the worst value can be the worst state from a valid historical sequence. For indicators such as mortality rate, pollution concentration, energy consumption, or emission intensity where lower values ​​are better, the best value is at the lower end of the range. Legitimate upper and lower bounds serve as additional hard constraints and do not become invalid due to standardization truncation.

[0023] The system performs six four-endpoint interval recursions for each secondary indicator to obtain the achievable range of the original value and the achievable range of the standardized score in 2030. It then aggregates these ranges to obtain the score ranges for each primary target endpoint and calculates the highest achievable overall index. If 82.00 falls between the current overall index and the highest achievable index, the effective target index remains 82.00; otherwise, an error is returned according to the aforementioned rules, or the highest achievable index is automatically adopted.

[0024] In the primary objective terminal space, the system finds an equilibrium anchor point and, for each primary objective with a valid secondary index, calculates a skewed path. For example, for primary objective 11, it first calculates the highest achievable score for this objective when the overall index equals 82.00, and then minimizes the perturbation to the equilibrium anchor point within a neighborhood no more than ε away from the highest score. This process is repeated for other primary objectives, ultimately generating K skewed paths and one equilibrium path.

[0025] During the path reconstruction phase, the system stores the forward reachability intervals from 2024 to 2030 for each secondary indicator and backpropagates from the 2030 terminal value to obtain the backward reachability interval. For each intermediate year, a value is selected from the intersection of the forward and backward intervals according to the smoothed progress. Secondary indicators belonging to the primary objective 11 that are emphasized adopt a progress bias of heavier progress at the beginning and lighter progress at the end, while secondary indicators not belonging to the primary objective adopt a progress bias of lighter progress at the beginning and heavier progress at the end, thus making the emphasized path identifiable from the balanced path on the time axis.

[0026] Recalculate all constraints for each path. Taking the focus path of primary objective 11 as an example, check whether the original values ​​of each secondary indicator from 2024 to 2030 are within the legal upper and lower bounds, and whether the rate of change in adjacent years is within […]. , Within the range, is the residual of the overall index after re-aggregation of terminal states less than δ? If all pass, the path is marked as valid and output; if the rate of change in a certain year exceeds the range, the index number, the year of violation and the constraint type are recorded, and the balanced path is automatically downgraded as the alternative output for this focus scenario.

[0027] The output includes: Requested target index 82.00, effective target index 82.00, target reachability status is reachable, planning year sequence from 2024 to 2030, validity markers for each scenario, annual trajectory of the overall index, annual trajectory of scores for each primary target, annual trajectory of original values ​​and rates of change for each secondary indicator, annual trajectory of coupling coordination degree, and verification report.

[0028] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An adaptive path planning method oriented towards urban sustainable development goals, characterized in that, Includes the following steps: S1. Obtain the hierarchical indicator system configuration, current status year, requested target index, target year, and parameters of each secondary indicator for the area to be planned; the parameters of the secondary indicators include the current value, best value, worst value, legal lower bound, legal upper bound, minimum annual change rate, and maximum annual change rate; S2. Based on the best and worst values, map the secondary indicators with different directions and dimensions to a unified score domain, and obtain the scores of each primary target and the current overall index according to the configured aggregation method. S3. Starting from the current value of the secondary indicator, propagate the reachable range of each secondary indicator year by year based on the annual minimum change rate, annual maximum change rate, legal lower bound, and legal upper bound to obtain the terminal reachable range corresponding to the target year. S4. Obtain the highest reachable value of the overall index based on the terminal reachable range. Output an input error when the requested target index is lower than the current overall index. Determine the highest reachable value as the valid target index when the requested target index is higher than the highest reachable value. Otherwise, determine the requested target index as the valid target index. S5. Under the constraint that the weighted sum of the scores of the first-level objectives equals the effective objective index and each score of the first-level objectives is within its respective reachable interval, the equilibrium anchor point is solved with the deviation of each score of the first-level objectives from the effective objective index as the objective function. S6. For each primary objective with effective secondary indicators, first solve for the maximum score of the primary objective under the same effective objective constraints as in step S5, then keep the score of the primary objective within the preset neighborhood of the maximum score, and solve for the focus endpoint with the goal of minimizing the disturbance of the relatively balanced anchor point. S7. Invert the equilibrium anchor point and each emphasis endpoint into the terminal value of the secondary indicator. Based on the intersection of the forward reachable interval and the backward reachable interval derived from the terminal value, select feasible secondary indicator values ​​year by year to form the equilibrium path and multiple emphasis paths. S8. Verify the numerical finiteness, legal upper and lower bounds, annual change rate, and overall index residual of the endpoint for each path, and output the first-level target path, second-level indicator path, overall index path, and coupling coordination degree path that have passed the verification. S9. If a path fails the verification, record the primary target number and the year of violation corresponding to the path, and automatically downgrade to use the feasible path from the previous iteration as the alternative output.

2. The adaptive path planning method for urban sustainable development goals as described in claim 1, characterized in that, The specific process of step S2 is as follows: S21. Map the secondary indicators to a unified score range. The calculation formula is as follows: ,in, is the standardized score of the raw value of the j-th secondary indicator; j is the index of the secondary indicator, j=1,2,…,N, and N is the total number of valid secondary indicators; This represents the original value of the j-th secondary indicator; This represents the best value of the j-th secondary indicator; Let be the worst value of the j-th secondary indicator, and Not equal to ; Π [0,100] For projection operators onto the closed interval [0, 100], results less than 0 are taken as 0, and results greater than 100 are taken as 100; for positive indices, > For negative indicators, < ; S22. The formula for calculating the score of the first-level objective is: ,in, The score is the score of the g-th primary objective; g is the primary objective number, g=1,2,…,K, and K is the total number of primary objectives with at least one valid secondary indicator; This is the set of valid secondary indicator indices belonging to the g-th primary objective; The non-negative aggregate weight of the j-th secondary indicator; primary objectives without effective secondary indicators are not included in the K, overall index, and optimization variable set, and are marked as having no data in the output; The current formula for calculating the overall index is: Where S is the current overall index; The normalized non-negative weights for the g-th first-level objective; each The sum is 1.

3. The adaptive path planning method for urban sustainable development goals as described in claim 2, characterized in that, In step S3, the j-th secondary index is propagated using a four-endpoint product propagation method in year t, let... , ,in and The candidate set consists of the minimum annual rate of change and the maximum annual rate of change. ,in, Let j be the candidate set of the four endpoints of the j-th secondary indicator in year t; and These represent the lower and upper ends of the reachable range of the j-th secondary indicator in year t-1, respectively. and These represent the lower and upper ends of the annual multiplicative factor for the j-th secondary indicator, respectively; t represents the year. The formula for calculating the reachable interval in year t is: ,in, and These represent the lower and upper ends of the reachable interval of the j-th secondary indicator in year t, respectively. and Let $\mathbf{j}$ be the legal lower bound and legal upper bound of the $j$-th secondary index, respectively; $\min(·)$ and $\max(·)$ are the minimum and maximum value operations, respectively; the initial condition is... , This represents the original value of the j-th secondary indicator in the current year.

4. The adaptive path planning method for urban sustainable development goals as described in claim 3, characterized in that, In step S4, the number of planning steps equals the difference between the target year and the current state year. The formula for calculating the effective target index is: Where T is the number of planning steps; The target year set by the user; The current state year; For effective target index; The target index for the user's request; This represents the highest achievable value of the overall index within the planning steps T. When T≤0, output "Target Year Error"; when the requested target index is lower than the current overall index, output "Target Index Lower than Current Overall Index"; where the current overall index is the overall index calculated from the current state.

5. The adaptive path planning method for urban sustainable development goals as described in claim 4, characterized in that, In step S5, the formula for calculating the equilibrium anchor point is: , Where B is the equilibrium anchor point; s is the first-level target terminal score vector; and These are the lower and upper bounds of the terminal reachability score for the g-th primary objective, respectively.

6. The adaptive path planning method for urban sustainable development goals as described in claim 5, characterized in that, In step S6, the k-th primary objective is solved using a two-stage lexicographical order approach. The first stage is: The first stage uses the same first-level objective reachability score boundary constraints and effective objective exponential equality constraints as step S5; the second stage is: , Where k is the index of the primary target currently being emphasized, k∈{1,2,…,K}; This represents the highest attainable score of the k-th primary objective under the effective objective index constraint. The terminal score for the k-th primary objective; Let be the stable endpoint vector of the extreme neighborhood of the k-th primary objective; The g-th first-order target component of the equalization anchor point; ε is the non-negative extreme value neighborhood tolerance; no tolerance is set in the second stage. A mandatory dominance constraint that is higher than the scores of all other first-level objectives.

7. The adaptive path planning method for urban sustainable development goals as described in claim 6, characterized in that: In step S6, the formula for calculating the kth emphasis endpoint of the adjustable separation degree is: ,in, This is the primary endpoint of the k-th primary objective; This is the path separation coefficient; since B and All are located on the same effective target hyperplane. Keep the overall index equal to .

8. The adaptive path planning method for urban sustainable development goals as described in claim 7, characterized in that: In step S7, for each secondary indicator, the reachable intervals for each year obtained by forward propagation from the current value are first saved, and then the backward reachable intervals for each year are obtained by backward propagation from the terminal value according to the annual rate of change. In the intersection of the forward reachable interval, backward reachable interval, and legal upper and lower bounds of the corresponding year, the annual value is selected according to the preset smooth progress; the secondary indicators to which the primary target belongs and the secondary indicators not to which the primary target belongs adopt opposite progress biases to enhance the identifiability of the focus path on the time axis.

9. The adaptive path planning method for urban sustainable development goals as described in claim 8, characterized in that, In step S8, the formulas for calculating the coupling degree, coordination comprehensive index, and coupling coordination degree are as follows: ,in, Coupling degree; Score for the g-th primary objective The dimensionless normalized value, And 0≤ ≤1; denoted as the comprehensive coordination index; D represents the coupling coordination degree.

Images