A multi-objective two-level optimization method for land use around rail transit stations

Abstract

The present invention aims at optimizing the land around rail transit stations, and establishes a multi-objective double-layer optimization method for land use around rail transit stations based on land use planning theory and method, traffic flow distribution theory and Pareto optimization theory. Specifically, the following steps are included: Step 100, inputting the traffic network and land use information; Step 101, establishing a multi-objective optimization model for land use around the station oriented to the development of rail transit for the upper-level problem; Step 102, establishing a traffic distribution model under the condition of user equilibrium in the lower level; Step 103, solving the multi-objective double-layer planning model; Step 104, the model outputs the land use optimization result. The present invention can not only conduct theoretical analysis on the implementation utility of the land use layout around rail transit stations, but also provide technical support for traffic planners on how to make reasonable decisions on the land use configuration around rail transit stations.

Description

Technical Field

[0001] The invention belongs to the field of traffic planning, and in particular relates to a multi-objective double-layer optimization method for land use around a rail transit station. Background Art

[0002] Most cities in my country have high population density, scarce land resources, and prominent problems of transportation supply and demand. These conflicts reflect the necessity of vigorously developing public transportation systems based on rail transit in my country's large cities. In order to meet the huge transportation demand, rail transit has become a transportation development strategy commonly adopted by large cities in my country. However, new transportation facilities will induce new transportation demand, leading to industrial agglomeration and population growth in cities, bringing greater traffic flow problems. To control urban transportation demand, we must first manage the sources of urban transportation demand so that the transportation demand generated by land use matches the transportation supply generated by roads.

[0003] Therefore, optimizing land use around rail transit development-oriented stations and coordinating the relationship between transportation systems and land use are of great significance for reducing urban traffic congestion and environmental pollution, improving the efficiency of public transportation and surrounding land use, and promoting sustainable urban development.

[0004] Maximizing rail transit passenger flow and minimizing the total carbon emissions of the road network are a pair of decision-making objectives that need to be balanced. The tendency and comprehensive optimization of the two will produce different optimization decisions. Taking the two as decision-making objectives to carry out land use optimization has great theoretical research significance and practical significance. The research on sustainable land use planning models around rail transit stations is still in the exploratory stage. Summary of the invention

[0005] The present invention aims to provide a multi-objective double-layer optimization method for land use around rail transit stations to solve the above problems.

[0006] The technical solution of the present invention is:

[0007] Step 100, inputting the traffic network and land use information; wherein the traffic network information within the influence range of the rail transit station includes the lane type, lane design traffic volume, speed and lane length; the land use information includes the land use type and building volume ratio;

[0008] Step 101, establishing a multi-objective optimization model for maximizing the rail transit passenger flow generated within the influence range of the upper rail transit station and minimizing the carbon emissions of cars on the road network around the station;

[0009] Step 102, under a given land use layout around a rail transit station, a traffic allocation model under a lower-level user equilibrium condition is established;

[0010] Step 103, solving a multi-objective two-level programming model, the upper layer uses a non-dominated sorting genetic algorithm (NSGA-II) to determine the land use type and building volume ratio planning scheme of each block, and the lower layer uses the Frank-Wolfe algorithm to solve the traffic flow distribution problem to obtain a Pareto optimal solution set;

[0011] Step 104: The model outputs land use optimization results.

[0012] Preferably, step 101 specifically includes the following steps:

[0013] Step 200 includes two objectives: objective 1: maximizing rail transit passenger flow, and objective 2: minimizing carbon emissions from cars on the road network;

[0014] Step 201, establishing constraints of the upper-level land use multi-objective optimization model.

[0015] Preferably, in step 200,

[0016] Objective 1 is specifically:

[0017]

[0018] Where: z1 is the total passenger flow of rail transit in the study area; i, j are the zones within the study area; k is the type of land use k∈[1,K]; The original unit for generating rail transit passenger flow for the k-type plot; The original unit for attracting rail transit passenger flow for the k-type plot; S i,j is the developable land area of ​​plot (i, j); is the plot ratio of the k-type land parcel (i, j); is the land use type of plot (i, j). When the land use type of plot (i, j) is k otherwise K L is the reduction coefficient of rail transit passenger flow with distance;

[0019] Objective 2 is specifically:

[0020] minz2=∑ a e(t a (V a ))

[0021] Section a has a flow rate of V a The carbon emissions generated under this condition can be expressed as:

[0022]

[0023] Where: z2 is the total carbon emissions of cars on the road network in the study area; e(t a (V a )) is the flow rate V a Carbon emissions generated by cars on the lower road section a; l a is the length of section a.

[0024] Preferably, the constraints of the upper-level problem in step 201 include the upper limit constraint of the plot ratio of key plots, the range constraint of the planned building plot ratio, the land compactness constraint, the land diversity constraint and the land allocation type constraint, which are as follows:

[0025] 1) Upper limit of plot ratio for key plots

[0026]

[0027] Where: is the upper limit of the plot ratio of the plot (i, j) calculated based on OD back-calculation; is the maximum traffic volume of the plot (i, j) obtained by OD reverse calculation; is the maximum traffic attraction of the plot (i, j) obtained by OD reverse calculation; p k is the original unit of traffic flow in the k-type plot; q k Attract the original unit for the traffic flow of the k-type plot;

[0028] 2) Planning floor area ratio range constraints

[0029]

[0030] Where: are the minimum and maximum building volume ratios of the land use type k in the planning of plot (i, j);

[0031] 3) Land compactness constraints

[0032] Pick

[0033]

[0034] Where: CCK is the land compactness index, is the current land compactness index, the smaller the value, the better the compactness; C k,k' is the quantitative value of the land use similarity between plot type k and plot type k'; i',j' is the adjacent plot (i',j') and plot (i,j)≠(i,'j');

[0035] 4) Land diversity constraints

[0036] When LM>0.8, the regional land layout is suitable for residents to walk. When LM<0.5, the regional land layout is not suitable for residents to walk. Take LM>0.5:

[0037] LM=∑ k (A k / A)ln(A k / A) / ln(K)

[0038]

[0039] A=∑ k A k

[0040]

[0041] Where: LM represents the land use diversity index; A is the total building area of ​​all land use types; A k is the total building area of ​​the kth land use type; K is the total land use type in the study area;

[0042] 5) Land allocation type constraints

[0043] Each parcel is assigned a land use type:

[0044]

[0045] Preferably, step 102 specifically includes the following steps:

[0046] Step 301, generate car traffic volume:

[0047] Q P =∑ i ∑ j S i,j ·R i,j ·p k

[0048] Q A =∑ i ∑ j S i,j ·R i,j ·q k

[0049] Where: Q P is the traffic volume; Q A To attract traffic;

[0050] Step 302, traffic distribution; the traffic flow from r to destination s is:

[0051]

[0052] Where: Qr,s is the traffic flow between OD pairs (r, s); r D is the traffic volume generated by the starting point r community; s is the traffic volume attracted by the destination s community; S' is the set of destinations s, s * ∈S';P(l r,s ) is the distance factor of the gravity model;

[0053] Step 303, traffic allocation under user equilibrium conditions:

[0054]

[0055] Where: t a (V a ) is at V a The travel time function of section a under traffic flow; f l rs is the flow on the lth path connecting the OD pair (r, s); is the upper limit of the plot ratio at the starting point of OD reverse calculation, R r The current plot ratio of the starting point; The upper limit of the destination plot ratio is calculated by OD, R s is the current plot ratio of the destination; is the path / segment association relationship. If segment a is on the lth path connecting the OD pair (r, s), then otherwise

[0056] Preferably, step 103 specifically includes the following steps:

[0057] Step 400, using NSGA-II algorithm to solve the upper multi-objective optimization, determine the land use type [x1, x2, ..., x n ] and building volume ratio [R1,R2,...,R n ];

[0058] Specifically, the land use type and volume ratio codes are converted into binary variables. Secondly, the road network data is loaded, the traffic generation and traffic attraction generated by each individual plot are calculated, and the traffic is input into the lower model. The speed of the road section with the highest speed limit is set to the free flow speed, and the Frank-Wolfe algorithm is used to distribute UE traffic to obtain the balanced flow of the road section. Finally, the upper model is returned to calculate the rail transit passenger flow and the carbon emissions of cars in the road network in the study area.

[0059] Step 401, the FW algorithm solves the lower-level traffic allocation problem; specifically, first, the traffic mode selection only includes rail transit travel and motor vehicle travel, the travel primitive units of the two travel modes are given, the land primitive unit method is used to calculate the traffic demand, and the traffic mode division step is omitted; secondly, the gravity model is used for traffic distribution prediction; finally, the Floyd algorithm is used to obtain the shortest path cost and the corresponding path, and the Frank-Wolfe algorithm is used to call the all-or-nothing allocation algorithm for UE traffic allocation.

[0060] Preferably, step 104 specifically includes the following steps:

[0061] First, construct a normalized initial matrix: There are n solutions in the first level of the Pareto optimal solution set, and each solution has two target indicators. Then the original data matrix is ​​constructed as S = S n×2 ;

[0062] Secondly, construct a weighted normalization matrix: normalize the attributes in the original data of the Pareto optimal solution set, so that each column element of the original data matrix is ​​divided by the norm of the current column vector to obtain a normalized matrix T = T n×2 ;

[0063] Finally, determine the best and worst solutions: calculate the closeness index C between each solution in the Pareto optimal solution set and the optimal level n .

[0064] The beneficial effects of the present invention are:

[0065] The present invention can not only conduct theoretical analysis on the implementation effectiveness of land use layout around rail transit stations, but also provide technical support for transportation planners on how to make reasonable decisions on land use configuration around rail transit stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A flowchart of a multi-objective dual-layer optimization method for land use around a rail transit station provided by an embodiment of the present invention;

[0067] Figure 2 A flow chart of a multi-objective optimization model for land use around a rail transit station established by a multi-objective dual-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0068] Figure 3 A flow chart of traffic allocation under the lower user equilibrium condition of a multi-objective double-layer optimization method for land use around a rail transit station provided by an embodiment of the present invention;

[0069] Figure 4A schematic diagram of land optimization configuration around a rail transit station according to a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0070] Figure 5 A flow chart of a NSGA-II solution algorithm for multi-objective bi-level planning of land use in a multi-objective bi-level optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0071] Figure 6 A schematic diagram of the assumed current land use volume ratio and land use type distribution of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0072] Figure 7 A Pareto optimal frontier surface diagram of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0073] Figure 8 A maximum value variation trend diagram of objective 1 of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0074] Fig. 9 A graph showing a change trend of the minimum value of objective 2 of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0075] Fig.10 A Pareto number variation trend diagram of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0076] Fig.11 A distribution result diagram of land use optimization that only considers ecological benefits in a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0077] Fig.12 An ecological priority land use optimization distribution result diagram of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0078] Fig.13 A distribution result diagram of land use optimization with equal emphasis on ecology and economy of a multi-objective dual-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0079] Fig.14 An economic priority land use optimization distribution result diagram of a multi-objective double-layer optimization method for land use around a rail transit station provided in an embodiment of the present invention;

[0080] Fig.15A land utilization optimization distribution result diagram that only considers economic benefits of a multi-objective double-layer optimization method for land utilization around a rail transit station provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0081] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. The implementation methods of the present invention are not limited thereto.

[0082] Figure 1 1 is a flow chart of the design scheme in the embodiment of the present invention. Figure 1 As shown, the multi-objective dual-layer optimization method for land use around a station guided by rail transit in an embodiment of the present invention includes the following steps:

[0083] Step 100: input the traffic network and land use information. The traffic network information within the influence range of the rail transit station includes lane type, lane design traffic volume, speed and lane length, etc.; land use information includes land use type and building volume ratio;

[0084] Step 101, establishing a multi-objective optimization model for maximizing the rail transit passenger flow generated within the influence range of the upper rail transit station and minimizing the carbon emissions of cars on the road network around the station;

[0085] Step 102, under a given land use layout around a rail transit station, a traffic allocation model under a lower-level user equilibrium condition is established;

[0086] Step 103, solving a multi-objective two-level programming model, wherein the upper level uses the NSGA-II genetic algorithm to determine the land use type and building volume ratio planning scheme of each block, and the lower level uses the Frank-Wolfe algorithm to solve the problem of traffic flow distribution;

[0087] Step 104: The model outputs the land use optimization result. In order to obtain the best solution, multiple weight combinations are set, and the TOPSIS method is used to evaluate and rank the solutions in the Pareto optimal solution set, and the most reasonable land use optimization solution around the station oriented to the development of rail transit is selected.

[0088] Below, the important steps in the above process are described in detail:

[0089] 1) Regarding step 101: Figure 2 This is a flow chart of a multi-objective optimization model for land use around stations guided by upper rail transit development established in an embodiment of the present invention. Specifically, it includes the following steps:

[0090] Step 200: A multi-objective two-level planning model for land use oriented to rail transit development is constructed. The generated rail transit demand is estimated using the original land unit method based on the traffic occurrence rate, building volume ratio and rail transit sharing rate of each plot.

[0091] Objective 1: Maximize the rail transit passenger flow generated in the rail transit station influence area

[0092]

[0093] Where: z1 is the total passenger flow of rail transit in the study area; i, j are the zones within the study area; k is the type of land use k∈[1,K]; The original unit for generating rail transit passenger flow for the k-type plot; The original unit for attracting rail transit passenger flow for the k-type plot; S i,j is the developable land area of ​​plot (i, j); is the plot ratio of the k-type land parcel (i, j); is the land use type of plot (i, j). When the land use type of plot (i, j) is k otherwise K L It is the reduction coefficient of rail transit passenger flow with distance.

[0094] Objective 2: Minimizing car carbon emissions from the road network

[0095] minz2=∑ a e(t a (V a ))

[0096] Section a has a flow rate of V a The carbon emissions generated under this condition can be expressed as:

[0097]

[0098] Where: z2 is the total carbon emissions of cars on the road network in the study area; e(t a (V a )) is the flow rate V a Carbon emissions generated by cars on the lower road section a; l a is the length of section a;

[0099] Step 201, establish the constraints of the upper-level land use multi-objective optimization model. The constraints of the upper-level problem include the upper limit constraint of the key plot volume ratio, the range constraint of the planned building volume ratio, the land compactness constraint, the land diversity constraint and the land allocation type constraint, which are as follows:

[0100] ① Upper limit of plot ratio for key plots

[0101]

[0102] Where: is the upper limit of the plot ratio of the plot (i, j) calculated based on OD back-calculation; is the maximum traffic volume of the plot (i, j) obtained by OD reverse calculation; is the maximum traffic attraction of the plot (i, j) obtained by OD reverse calculation; p k is the original unit of traffic flow in the k-type plot; q k Attract the original unit for the traffic flow of type k plots.

[0103] ② Planning floor area ratio range constraints

[0104]

[0105] Where: are the minimum and maximum building volume ratios of land use type k in the planning of plot (i, j).

[0106] ③Land compactness constraints

[0107] The more similar the land use of the units allocated around the plot, the higher the density and the convenience of interaction between activities. The land use type improves this convenience. If all units are developed into the same land use type, the land compactness is the best. At the same time, the degree of land use conflict between adjacent units needs to be reduced to ensure the quality of life of residents.

[0108]

[0109] Where: CCK is the land compactness index, is the current land compactness index, the smaller the value, the better the compactness; C k,k' is the quantitative value of the land use similarity between plot type k and plot type k'; i',j' means that plot (i',j') is adjacent to plot (i,j)≠(i,'j').

[0110] ④Land diversity constraints

[0111] When LM>0.8, the regional land layout is suitable for residents to walk. When LM<0.5, the regional land layout is not suitable for residents to walk. Take LM>0.5.

[0112] LM=∑ k (A k / A)ln(A k / A) / ln(K)

[0113]

[0114] A=∑ k A k

[0115]

[0116] Where: LM represents the land use diversity index; A is the total building area of ​​all land use types; A k is the total building area of ​​the kth land use type; K is the total land use type in the study area.

[0117] ⑤ Constraints on land allocation type

[0118] Each parcel within the study area can only be assigned one land use type.

[0119]

[0120]

[0121] 2) Regarding step 102: Figure 3 It is a traffic allocation flow chart under the lower-level user equilibrium condition established under a given land use layout around a rail transit station in an embodiment of the present invention. Figure 4 This is a schematic diagram of the optimized land configuration around rail transit stations. It specifically includes the following steps:

[0122] Step 301, generate car traffic volume.

[0123] Q P =∑ i ∑ j S i,j ·R i,j ·p k

[0124] Q A =∑ i ∑ j S i,j ·R i,j ·q k

[0125] Where: Q P is the traffic volume; Q A To attract traffic.

[0126] Step 302, traffic distribution. Assumptions: 1) The flow of commuters from the origin area r to the destination area s is proportional to the traffic volume at the destination; 2) The flow of commuters depends on the distance l between the origin and the destination. r,s , and a distance factor P(l r,s). Using these assumptions, the traffic flow from r to destination s is:

[0127]

[0128] Where: Q r,s is the traffic flow between OD pair (r, s); r is the traffic volume generated by the starting point r community; D s is the traffic volume attracted by the destination s community; S' is the set of destinations s, s * ∈S';P(l r,s ) is the distance factor of the gravity model.

[0129] Step 303: Traffic allocation under user equilibrium conditions.

[0130]

[0131] Where: t a (V a ) is at V a The travel time function of section a under flow; f l rs is the flow on the lth path connecting the OD pair (r, s); is the upper limit of the plot ratio at the starting point of OD reverse calculation, R r The current plot ratio of the starting point; The upper limit of the destination plot ratio is calculated by OD, R s is the current plot ratio of the destination; is the path / segment association relationship. If segment a is on the lth path connecting the OD pair (r, s), then otherwise

[0132] 3) Regarding step 103, solve the multi-objective bi-level programming model. Figure 5 This is the flow chart of the NSGA-II solution algorithm for multi-objective bi-level land use programming. It includes the following steps:

[0133] Step 400: First, the upper layer problem is a multi-objective optimization problem of land use. The NSGA-II algorithm is used to solve the upper layer multi-objective optimization problem and determine the land use type [x1, x2, ..., x n ] and building volume ratio [R1,R2,...,R n]. Convert the land use type and floor area ratio codes into binary variables. Secondly, load the road network data, calculate the traffic generation and traffic attraction generated by each individual plot, and input them into the lower model. Set the free flow speed when the speed limit of the road section is the highest, use the Frank-Wolfe algorithm to distribute UE traffic, and obtain the balanced flow of the road section. Finally, return to the upper model to calculate the rail transit passenger flow and road network car carbon emissions in the study area, find the Pareto optimal solution set of the optimization problem, and generate multiple planning schemes. Use the TOPSIS method to evaluate and rank the solutions in the Pareto optimal solution set in the land use optimization results to determine the best land use configuration plan.

[0134] Step 401, use the Frank-Wolfe algorithm to solve the lower-level traffic distribution problem. First, the traffic mode selection only includes rail transit travel and motor vehicle travel. The original travel units of the two travel modes are given, and the land original unit method is used to calculate the traffic demand, so the traffic mode division step is omitted. Secondly, the gravity model is used to predict traffic distribution. Finally, the Floyd algorithm is used to obtain the shortest path cost and the corresponding path, and the Frank-Wolfe algorithm is used to call the all-or-nothing allocation algorithm for UE traffic allocation. Use the Frank-Wolfe algorithm to solve the Beckmann model. For the Beckmann model, the starting point of the known iteration The next iteration direction is:

[0135]

[0136] It can be seen The objective function of formula (2-9) can be transformed into Therefore, the linear programming problem can be expressed as:

[0137]

[0138] Among them, y a is the feasible road flow at the nth iteration; is the feasible path flow at the nth iteration. According to the all-or-nothing distribution method, ya is solved, and the fastest descent direction of the objective function is (X n -Y n ). The next problem is to determine the optimal iteration step size λ in the direction of steepest descent in the nth iteration:

[0139]

[0140] make get There is only one variable λ in the equation, which is calculated by dichotomy. Therefore, it can be obtained by Calculate the starting point for the next iteration

[0141] 4) Regarding step 104, the model outputs the land use optimization result. In order to obtain the best solution, multiple weight combinations are set in the end, and the TOPSIS method is used to evaluate and rank the solutions in the Pareto optimal solution set, and the most reasonable land use optimization solution around the station oriented to the development of rail transit is selected. Specifically, the following steps are included:

[0142] First, construct a normalized initial matrix. There are n solutions in the first level of the Pareto optimal solution set, and each solution has two target indicators. Then the original data matrix is ​​constructed as S = S n×2 .use Secondly, a weighted normalization matrix is ​​constructed to normalize the attributes in the original data of the Pareto optimal solution set, so that each column element of the original data matrix is ​​divided by the norm of the current column vector to obtain a normalized matrix T = T n×2 Finally, determine the best solution and the worst solution: The best solution T + It consists of the maximum value among the elements in each row of T. The worst solution T - It consists of the maximum value of each column element in T Calculate the closeness index between each solution in the Pareto optimal solution set and the optimal level C n The larger the index, the closer it is to the optimal level. is the closeness index between the nth solution and the worst solution, is the closeness index between the nth solution and the optimal solution, w j is the weight of the j-th target.

[0143] The present invention is further described below by taking a numerical experiment as an example:

[0144] Assume that the side length of each plot in the study area is 200 meters, plot number 13 is fixed as the rail transit station development area, and the current land use volume ratio and land use type distribution are as follows: Figure 6As shown. Assume that 40% of the land in the study area is used for public facilities such as roads, and 20% of the remaining 60% of the land is used for green land within the community. Related parameter settings: ① The morning peak traffic generation rate under different land use types is shown in Table 1, and the morning peak rail transit passenger flow generation rate is shown in Table 2; ② The reduction coefficient of the rail transit share rate with distance is shown in Table 3; ③ The upper and lower limits of various land types and circle volume ratios are shown in Table 4; ④ The road capacity is shown in Table 5; ⑤ The parameters of the travel distribution gravity model are shown in Table 6; ⑥ The degree of conflict between the two assumed land types is shown in Table 7. Use MATLAB to write the NSGA-Ⅱ algorithm. In this experiment, the population size is selected as 200, the maximum evolutionary generation is 1000, the crossover probability is set to 0.7, and the mutation probability is 0.05. The Pareto solution set is solved, and each solution in the solution set corresponds to a land use optimization plan. In order to obtain the best solution, multiple weight combinations are set in the end, and the TOPSIS method is used to evaluate and rank the solutions in the Pareto optimal solution set.

[0145] Table 1 Traffic generation rate and attraction rate of each land use type under unit development intensity during the morning peak

[0146]

[0147] Table 2 Rail transit passenger flow generation rate under each land use type unit development intensity during the morning peak

[0148]

[0149] Table 3 Reduction coefficient of rail transit share rate with distance

[0150]

[0151] Table 4 Land use type and range of plot ratio values

[0152]

[0153] Table 5 Road Capacity

[0154]

[0155] Table 6 Parameters of the travel distribution gravity model

[0156]

[0157] Table 7 Conflict degree between land use types

[0158]

[0159] In order to explore the optimal land use type and building volume ratio of the plots in the study area without changing the existing traffic network, a numerical experiment was set up. NSGA-II was used to solve the multi-objective double-layer model of land use and generate the Pareto frontier surface, such as Figure 7 Each solution in the Pareto solution set corresponds to a land use plan, and the generated partial Pareto optimal solution set is shown in Table 8.

[0160] Table 8 Part of the Pareto optimal solution set generated

[0161]

[0162] The ecological benefit objective functions of rail transit operation efficiency and regional carbon emissions are processed to optimize different objective functions in their respective directions as much as possible. From the iteration trend of the maximum value of rail transit passenger flow, the iteration trend of the minimum value of road network carbon emissions, and the iteration trend of the Pareto number, it can be seen that the algorithm used in this paper can converge. The minimum value of each target per iteration can be reduced by increasing the number of iterations and finally reach a stable state, such as Figure 8 , Fig. 9 and Fig.10 As shown in Table 9. Since the final optimization result needs to be selected according to the decision maker's preference, in order to avoid the weight selection being too subjective, this paper uses the TOPSIS method to perform decision analysis on the multi-objective optimization solution set. The land use optimization scenario model is shown in Table 9. The solutions of order 1 and the current situation solutions in the Pareto optimal solution set under different scenarios are shown in Table 10.

[0163] Scenario 1: Considering only ecological benefits, the land use optimization distribution results are as follows Fig.11 As shown;

[0164] Scenario 2: Ecological priority, land use optimization distribution results are as follows Fig.12 As shown;

[0165] Scenario 3: Ecology and economy are equally important, and the results of land use optimization distribution are as follows Fig.13 As shown;

[0166] Scenario 4: Economic priority, land use optimization distribution results are as follows Fig.14 As shown;

[0167] Scenario 5: Considering only economic benefits, the land use optimization distribution results are as follows Fig.15 shown.

[0168] Table 9 Land use optimization scenario model

[0169]

[0170] Table 10 Solutions of order 1 and current status in the Pareto optimal solution set under different scenarios

[0171]

[0172] The above numerical example results demonstrate the effectiveness and practicality of the model constructed by the present invention, indicating that after optimizing the building volume ratio and land use allocation type in the integrated optimization model of transportation and land use, this paper can find a land use optimization solution to increase rail transit passenger flow and reduce the total carbon emissions of small cars in the road network.

[0173] The above description is only a preferred embodiment of the present invention. Any changes and modifications made according to the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A multi-objective double-layer optimization method for land use around rail transit stations, characterized in that: The following steps are involved: Step 100, inputting the traffic network and land use information; wherein the traffic network information within the influence range of the rail transit station includes the lane type, lane design traffic volume, speed and lane length; the land use information includes the land use type and building volume ratio; Step 101, establishing a multi-objective optimization model for maximizing the rail transit passenger flow generated within the influence range of the upper rail transit station and minimizing the carbon emissions of cars on the road network around the station; Step 102, under a given land use layout around a rail transit station, a traffic allocation model under a lower-level user equilibrium condition is established; Step 103, solving a multi-objective two-level programming model, the upper layer uses a non-dominated sorting genetic algorithm (NSGA-II) to determine the land use type and building volume ratio planning scheme of each block, and the lower layer uses the Frank-Wolfe algorithm to solve the traffic flow distribution problem to obtain a Pareto optimal solution set; Step 104, the model outputs land use optimization results; Wherein, step 101 specifically includes the following steps: Step 200 includes two objectives: objective 1: maximizing rail transit passenger flow, and objective 2: minimizing carbon emissions from cars on the road network; Step 201, establishing constraints of the upper-level land use multi-objective optimization model; Among them, in step 200, Objective 1 is specifically: Where: z1 is the total passenger flow of rail transit in the study area; i, j are the zones within the study area; k is the type of land use k∈[1,K]; The original unit for generating rail transit passenger flow for the k-type plot; The original unit for attracting rail transit passenger flow for the k-type plot; S i,j is the developable land area of ​​plot (i, j); is the plot ratio of the k-type land parcel (i, j); is the land use type of plot (i, j). When the land use type of plot (i, j) is k otherwise K L is the reduction coefficient of rail transit passenger flow with distance; Objective 2 is specifically: minz2=∑ a e(t a (V a )) Section a has a flow rate of V a The carbon emissions generated under this condition can be expressed as: Where: z2 is the total carbon emissions of cars on the road network in the study area; e(t a (V a )) is the flow rate V a Carbon emissions generated by cars on the lower road section a; l a is the length of section a; The constraints of the upper-level problem in step 201 include the upper limit constraint of the plot ratio of key plots, the range constraint of the planned building plot ratio, the land compactness constraint, the land diversity constraint, and the land allocation type constraint, which are as follows: 1) Upper limit of plot ratio for key plots Where: is the upper limit of the plot ratio of the plot (i, j) calculated based on OD back-calculation; is the maximum traffic volume of the plot (i, j) obtained by OD reverse calculation; is the maximum traffic attraction of the plot (i, j) obtained by OD reverse calculation; p k is the original unit of traffic flow in the k-type plot; q k Attract the original unit for the traffic flow of the k-type plot; 2) Planning floor area ratio range constraints Where: are the minimum and maximum building volume ratios of the land use type k in the planning of plot (i, j); 3) Land compactness constraints Pick Where: CCK is the land compactness index, is the current land compactness index, the smaller the value, the better the compactness; C k,k' is the quantitative value of the land use similarity between plot type k and plot type k'; i',j' is the adjacent plot (i',j') and plot (i,j)≠(i,'j'); 4) Land diversity constraints When LM>0.8, the regional land layout is suitable for residents to walk. When LM<0.5, the regional land layout is not suitable for residents to walk. Take LM>0.5: LM=∑ k (A k / A)ln(A k / A) / ln(K) A=∑ k A k Where: LM represents the land use diversity index; A is the total building area of ​​all land use types; A k is the total building area of ​​the kth land use type; K is the total land use type in the study area; 5) Land allocation type constraints Each parcel is assigned a land use type:

2. A multi-objective double-layer optimization method for land use around rail transit stations according to claim 1, characterized in that: Step 102 specifically includes the following steps: Step 301, generate car traffic volume: Q P =∑ i ∑ j S i,j ·R i,j ·p k Q A =∑ i ∑ j S i,j ·R i,j ·q k Where: Q P is the traffic volume; Q A To attract traffic; Step 302, traffic distribution; the traffic flow from r to destination s is: Where: Q r,s is the traffic flow between OD pair (r, s); r is the traffic volume generated by the starting point r community; D s is the traffic volume attracted by the destination s community; S' is the set of destinations s, s * ∈S';P(l r,s ) is the distance factor of the gravity model; Step 303, traffic allocation under user equilibrium conditions: Where: t a (V a ) is at V a The travel time function of section a under traffic flow; f l rs is the flow on the lth path connecting the OD pair (r, s); is the upper limit of the plot ratio at the starting point of OD reverse calculation, R r The current plot ratio of the starting point; The upper limit of the destination plot ratio is calculated by OD, R s is the current plot ratio of the destination; is the path / segment association relationship. If segment a is on the lth path connecting the OD pair (r, s), then otherwise 3. The multi-objective double-layer optimization method for land use around rail transit stations according to claim 1 is characterized in that: Step 103 specifically includes the following steps: Step 400, using NSGA-II algorithm to solve the upper multi-objective optimization, determine the land use type [x1, x2, ..., x n ] and building volume ratio [R1,R2,...,R n ]; Specifically, the land use type and volume ratio codes are converted into binary variables. Secondly, the road network data is loaded, the traffic generation and traffic attraction generated by each individual plot are calculated, and the traffic is input into the lower model. The speed of the road section with the highest speed limit is set to the free flow speed, and the Frank-Wolfe algorithm is used to distribute UE traffic to obtain the balanced flow of the road section. Finally, the upper model is returned to calculate the rail transit passenger flow and the carbon emissions of cars in the road network in the study area. Step 401, the FW algorithm solves the lower-level traffic allocation problem; specifically, first, the traffic mode selection only includes rail transit travel and motor vehicle travel, the travel primitive units of the two travel modes are given, the land primitive unit method is used to calculate the traffic demand, and the traffic mode division step is omitted; secondly, the gravity model is used for traffic distribution prediction; finally, the Floyd algorithm is used to obtain the shortest path cost and the corresponding path, and the Frank-Wolfe algorithm is used to call the all-or-nothing allocation algorithm for UE traffic allocation.

4. The multi-objective double-layer optimization method for land use around rail transit stations according to claim 1 is characterized in that: Step 104 specifically includes the following steps: First, construct a normalized initial matrix: There are n solutions in the first level of the Pareto optimal solution set, and each solution has two target indicators. Then the original data matrix is ​​constructed as S = S n×2 ; Secondly, construct a weighted normalization matrix: normalize the attributes in the original data of the Pareto optimal solution set, so that each column element of the original data matrix is ​​divided by the norm of the current column vector to obtain a normalized matrix T = T n×2 ; Finally, determine the best and worst solutions: calculate the closeness index C between each solution in the Pareto optimal solution set and the optimal level n .

Images