Dynamic cross-layer path planning method based on multiple constraint conditions

Abstract

The invention discloses a multi-constraint-condition-based dynamic cross-layer path planning method, which belongs to the technical field of indoor navigation, and comprises the following steps of: constructing a real-time constraint set containing a plurality of quantitative constraint conditions through multi-source data fusion, multi-constraint comprehensive decision and dynamic weight updating, continuously updating a'dynamic weighted multilayer topological graph 'representing an indoor space, and performing multi-source data fusion, multi-constraint comprehensive decision and dynamic weight updating. An optimal cross-layer navigation path is generated and dynamically adjusted in real time, and search efficiency and decision quality are considered based on a two-stage optimization processing mechanism (edge weight real-time updating and path evaluation screening), so that path planning can intelligently adapt to an indoor environment which changes in real time, and individual requirements of users are fully met; the critical defect that a static path library cannot adapt to a dynamic environment in the prior art is effectively overcome, and the accuracy, adaptability and user experience of path planning in a complex multi-layer indoor environment are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of indoor navigation technology, and more specifically, to a dynamic cross-layer path planning method based on multiple constraints. Background Technology

[0002] Indoor navigation technology, especially path planning across floors, is key to improving visitor experience and operational efficiency in large and complex buildings (such as shopping malls, airports, and hospitals). Traditional cross-floor path planning methods are usually based on complete indoor road network data and use classic graph search algorithms for real-time calculation. Although this method can guarantee finding feasible paths, it has high computational complexity when dealing with large-scale, multi-floor indoor maps, posing a challenge to the computing power and battery life of mobile terminals.

[0003] An improvement scheme is proposed in the prior art, such as the "Implementation Method of Cross-Floor Path Calculation Based on Indoor Map" disclosed in patent number CN104280028B. The core idea of ​​this patent is the "pre-calculation-retrieval" mode, which decomposes the cross-floor path planning into two parts: 1) Real-time calculation of the horizontal path from the starting point to the cross-floor node of the current floor (such as the elevator entrance) and the cross-floor node of the target floor to the destination; 2) The cross-floor part in the middle is obtained directly by querying a pre-calculated and stored static path library, and the paths obtained before and after are spliced ​​into a complete result.

[0004] While the above-mentioned processing methods effectively reduce real-time computation, improve response speed, and reduce terminal power consumption, they still have certain inherent limitations:

[0005] For example, static path libraries cannot adapt to dynamic environments: the cross-floor path libraries they rely on are pre-calculated and stored in a fixed manner, and cannot perceive and respond to real-time changes in the indoor environment.

[0006] If the planning dimension is too narrow, it will lack personalization and comprehensive optimization: the path pre-generation and retrieval process of the solution mainly (or even only) considers geometric distance or basic connectivity, and cannot incorporate multi-dimensional constraints. It cannot make comprehensive decisions based on the user's real-time preferences (such as barrier-free access and avoiding congestion), the real-time status of cross-floor facilities (such as elevator waiting time), dynamic pedestrian density or temporary building management rules (such as area restrictions), and therefore cannot provide a truly "optimal" or "most suitable" personalized path.

[0007] If decision-making capabilities are limited: once a path is generated, it cannot be dynamically adjusted according to real-time changes in the indoor environment (such as pedestrian density and equipment status), leading to planning failure or a decline in experience and insufficient real-time performance.

[0008] Therefore, in view of the actual technical defects, a dynamic cross-layer path planning method based on multiple constraints is proposed. Summary of the Invention

[0009] The purpose of this invention is to solve the problems of inflexible cross-level path planning, single constraints, and lack of real-time dynamic response. It provides a dynamic cross-level path planning method based on multiple constraints, which is applicable to complex multi-level indoor environments. It can integrate multi-dimensional constraints and dynamically generate the optimal cross-level path in real time, realizing dynamic environmental adaptation, comprehensive trade-off of multiple constraints, and personalized service.

[0010] The objective of this invention can be achieved through the following technical solution: a dynamic cross-layer path planning method based on multiple constraints, comprising the following steps:

[0011] S1. Receive user route planning request: Receive the user's input of the starting point, destination, and route selection preferences;

[0012] S2. Cross-layer data acquisition and fusion: Acquire multi-source data of the target space, including static geographic data, dynamic environmental data, user context data and building rule management data, and align and fuse the above four types of data in time and space to form unified spatial environment status data;

[0013] S3. Construction of multi-layer topology graph:

[0014] S31. Based on static geographic data, the indoor multi-level space is abstracted into an initial multi-level topology graph containing nodes and connecting edges of each floor, and each connecting edge is assigned an initial weight.

[0015] S32. Based on path selection preferences and spatial environment status data, extract and construct a real-time constraint set containing multiple quantitative constraints.

[0016] S33. The constraints that can be assigned to the connecting edges in the real-time constraint set are quantified by a preset edge weight update model, and the comprehensive dynamic weight of each edge in the initial multi-layer topology graph is calculated and updated to obtain a dynamically weighted multi-layer topology graph.

[0017] S4. Dynamic cross-layer multi-path generation:

[0018] S41. On a dynamically weighted multi-layer topology graph, run a path search algorithm to generate multiple candidate paths from the starting point to the ending point.

[0019] S42. Based on the real-time constraint set, multiple candidate paths are evaluated and screened as a whole through a pre-set path evaluation model, and the optimal dynamic cross-layer planning path that meets multiple constraints is output.

[0020] Furthermore, in S2, static geographic data includes data from BIM or interior maps, including floor plans, room / corridor outlines, and locations of fixtures (staircases, elevator shafts).

[0021] Dynamic environmental data includes data from sensor networks, such as the real-time operating status of the elevator, passenger density inside the car, passenger density in the waiting area, and passenger density in key areas (corridors, escalator entrances and exits).

[0022] User context data includes the user's real-time location coordinates, movement speed, and historical route selection preferences.

[0023] Rule management data includes information from building management systems, such as temporary closure notices for areas, access permission lists for specific access control systems, and route control information resulting from large-scale events.

[0024] Furthermore, in S31, key nodes include cross-floor facility nodes, representing path decision points such as corridor intersections, room entrances and exits, stairwells, elevator lobbies, and other key locations. Edges include horizontal movement edges and cross-floor virtual edges connecting corresponding cross-floor facility nodes on different floors, representing passable paths between key nodes.

[0025] Furthermore, in S32, based on path selection preferences and spatial environment status data, multi-dimensional constraints are extracted using a constraint building engine or constraint extraction logic to form a real-time constraint set.

[0026] The real-time constraint set includes at least two of the following types of constraints:

[0027] Edge-determinable constraints: Constraints that can be quantified and assigned to a single edge, including distance constraints based on the physical length of the path, time constraints based on real-time traffic conditions, comfort constraints based on the density of people in the area, and rule constraints set by the building management.

[0028] Path-level constraints: Constraints that need to be statistically analyzed or evaluated at the complete path level, including waiting and transfer constraints based on the real-time status of cross-level facilities, path type consistency constraints based on user preferences, and total path energy consumption constraints.

[0029] Furthermore, in step S33, the edge weight update model calculates a comprehensive dynamic weight value for each connecting edge based on the edgeizability constraint; in step S42, the path evaluation model generates an evaluation dimension score for each candidate path based on the edgeizability constraint and the path-level constraint, and selects the candidate path with the highest score as the dynamic cross-layer planning path.

[0030] The dynamic cross-layer path planning method based on multiple constraints also includes: S5, dynamic path replanning: outputs the dynamic cross-layer planned path and continuously monitors it during navigation, updates the real-time constraint set, performs dynamic replanning based on the preset real-time constraint triggering mechanism, determines whether the preset triggering conditions are met, and if so, re-executes S32-S42 to obtain the current optimal navigation path.

[0031] Furthermore, the real-time constraint triggering mechanism preset in S5 includes, but is not limited to, the following triggering conditions:

[0032] Environmental state change trigger condition: The change value of dynamic environmental state parameters related to the current navigation path exceeds the preset parameter threshold;

[0033] User deviation trigger conditions: The deviation distance between the user's actual location and the planned path exceeds a preset distance threshold, or the difference between the user's actual travel speed and the planned assumed speed exceeds a preset speed threshold;

[0034] Periodic trigger condition: The time interval since the last planning reaches a preset time threshold;

[0035] Rule change triggering conditions: When changes affecting management data are detected in the building management system, such as changes to the current route;

[0036] Predictive triggering conditions: Based on dynamic environmental data / rule management data prediction models (machine learning models, event-driven models), predict environmental state changes that will affect the current path within a specific time period in the future.

[0037] Furthermore, different priority weights are assigned to different triggering conditions, and when multiple triggering conditions are met simultaneously, they are processed in order of priority.

[0038] Compared with the prior art, the advantages of this invention are:

[0039] 1. This invention constructs a real-time constraint set containing multiple quantitative constraints by fusing multi-source data. Based on multi-constraint comprehensive decision-making, it continuously updates a "dynamic weighted multi-layer topology map" representing the indoor space. The edge weights in the topology map are no longer fixed values, but are uniformly quantified and updated in real time based on user personalized preferences, real-time dynamic environment and management rule constraints. This generates and dynamically adjusts the optimal cross-layer navigation path, effectively solving the fatal defect of being unable to adapt to dynamic environments due to the use of static path libraries, and greatly improving the reliability and scenario applicability of navigation.

[0040] 2. This invention also adopts a two-stage optimization processing mechanism for path generation that combines "real-time edge weight update" and "path evaluation and screening". In the first stage, some optional constraints are combined with the edge weight update model to "compile" complex constraints into a "cost map" that can be searched efficiently. The core objective is to ensure the feasibility, efficiency and initial response to dynamic environments of path search. In the second stage, all real-time constraint sets are combined with the path evaluation model to perform in-depth, global and multi-objective trade-offs on the candidate path set. The core objective is to ensure that the final output path achieves optimal performance in terms of higher-order and more complex user needs and overall experience. Attached Figure Description

[0041] Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention;

[0042] Figure 2 This is a flowchart of the method in Embodiment 2 of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0044] Example 1: This invention discloses a dynamic cross-layer path planning method based on multiple constraints. Please refer to [link / reference]. Figure 1 It includes the following steps:

[0045] S1. Receive user route planning request: Receive the user's input of the starting point, destination, and route selection preferences;

[0046] S2. Cross-layer data acquisition and fusion: Acquire multi-source data of the target space, including static geographic data, dynamic environmental data, user context data and building rule management data, and align and fuse the above four types of data in time and space to form unified spatial environment status data;

[0047] S3. Construction of multi-layer topology graph:

[0048] S31. Based on static geographic data, the indoor multi-level space is abstracted into an initial multi-level topology graph containing nodes and connecting edges of each floor, and each connecting edge is assigned an initial weight.

[0049] S32. Based on path selection preferences and spatial environment status data, extract and construct a real-time constraint set containing multiple quantitative constraints.

[0050] S33. The constraints that can be assigned to the connecting edges in the real-time constraint set are quantified by a preset edge weight update model, and the comprehensive dynamic weight of each edge in the initial multi-layer topology graph is calculated and updated to obtain a dynamically weighted multi-layer topology graph.

[0051] S4. Dynamic cross-layer multi-path generation:

[0052] S41. On a dynamically weighted multi-layer topology graph, run a path search algorithm to generate multiple candidate paths from the starting point to the ending point.

[0053] S42. Based on the real-time constraint set, multiple candidate paths are evaluated and screened as a whole through a pre-set path evaluation model, and the optimal dynamic cross-layer planning path that meets multiple constraints is output.

[0054] In S1, the input boundaries of the planning task are clearly defined, with user subjective intentions (preferences) as the primary input, distinguishing the differentiated needs of different user groups and laying the foundation for subsequent personalized planning.

[0055] In S2, static geographic data includes BIM or interior maps, including floor plans, room / corridor outlines, and locations of fixtures (stairs, elevator shafts).

[0056] Dynamic environmental data includes data from sensor networks, such as the real-time operating status of the elevator, passenger density inside the car, passenger density in the waiting area, and passenger density in key areas (corridors, escalator entrances and exits).

[0057] User context data includes the user's real-time location coordinates, movement speed, and historical route selection preferences.

[0058] Rule management data includes information from building management systems, such as temporary closure notices for areas, access permission lists for specific access control systems, and route control information resulting from large-scale events.

[0059] This process, through spatiotemporal alignment and confidence fusion, transforms multi-source heterogeneous data into a unified, consistent, and reliable "snapshot of spatial environment status," eliminating data conflicts and data silos.

[0060] Specifically: Time alignment: All data is timestamped, and the system retrieves data from the same or most recent moment at fixed time intervals (e.g., per second) or in an event-driven manner;

[0061] Spatial alignment: Mapping the location of dynamic environmental data (such as sensor data) to the coordinate system of static geographic data. For example, each sensor's installation location has preset coordinates in a static map, and the state it detects corresponds to a specific area or facility in the map.

[0062] In S31, key nodes include cross-floor facility nodes, which represent path decision points, such as key locations like corridor intersections, room entrances and exits, stairwells, and elevator lobbies. Edges include horizontal movement edges and cross-floor virtual edges that connect corresponding cross-floor facility nodes on different floors, representing passable paths between key nodes.

[0063] In S32, based on path selection preferences and spatial environment status data, multi-dimensional constraints are extracted using a constraint building engine or constraint extraction logic to form a real-time constraint set.

[0064] The real-time constraint set includes at least two of the following types of constraints:

[0065] Edge-determinable constraints: Constraints that can be quantified and assigned to a single edge, including distance constraints based on the physical length of the path, time constraints based on real-time traffic conditions, comfort constraints based on the density of people in the area, and rule constraints set by the building management.

[0066] Path-level constraints: Constraints that need to be statistically analyzed or evaluated at the complete path level, including waiting and transfer constraints based on the real-time status of cross-level facilities, path type consistency constraints based on user preferences, and total path energy consumption constraints.

[0067] The process of constructing a real-time constraint set is essentially to parse, transform, and parameterize two types of input data to form a complete and quantified set of planning instructions.

[0068] In S33, the edge weight update model calculates a comprehensive dynamic weight value for each connected edge based on the edgeability constraint. It should be added that in the dynamically weighted multi-layer topology graph, the passage cost of a node is reflected by the weight change of the edge associated with it. For nodes that need to wait or be processed, the node cost is simulated by creating a virtual edge at the node or adjusting the weight of the adjacent edge.

[0069] In step S42, the path evaluation model generates an evaluation dimension score for each candidate path based on the edgeability constraint and the path-level constraint, and selects the candidate path with the highest score as the dynamic cross-layer planning path, which is the optimal navigation path.

[0070] A two-stage optimization process for path generation, combining real-time edge weight updates and path evaluation and filtering, is adopted. In the first stage, a weight fusion model is used to dynamically map some multi-dimensional constraints onto the basic elements (connecting edges) of the graph in real time, generating a "dynamic cost map" that reflects the current comprehensive travel cost. This map is responsible for "compiling" complex constraints into a searchable "cost map." The core objective is to ensure the feasibility, efficiency, and initial response to dynamic environments in path search.

[0071] In the second phase, an independent "path evaluation model" is introduced to perform a global, multi-objective comprehensive evaluation of the complete path for all real-time constraint sets. In particular, it can handle path-level constraints such as "number of transfers" and "path type" that cannot be decomposed into a single edge, enabling more refined screening that better reflects users' real preferences. It is responsible for performing in-depth, global, and multi-objective trade-offs on the candidate path set. The core objective is to ensure that the final output path achieves optimal performance in terms of higher-order and more complex user needs and overall experience.

[0072] Example 2: This dynamic cross-layer path planning method, please refer to... Figure 2It also includes: S5, Path Dynamic Replanning: Outputs a dynamic cross-layer planned path and continuously monitors it during navigation, updates the real-time constraint set, performs dynamic replanning based on the preset real-time constraint triggering mechanism, determines whether the preset triggering conditions are met, if so, re-executes S32-S42 to obtain the current optimal navigation path, if not, continues the current navigation path.

[0073] The real-time constraint triggering mechanism preset in S5 includes, but is not limited to, the following triggering conditions:

[0074] Environmental state change trigger condition: The change value of dynamic environmental state parameters related to the current navigation path exceeds the preset parameter threshold;

[0075] User deviation trigger conditions: The deviation distance between the user's actual location and the planned path exceeds a preset distance threshold, or the difference between the user's actual travel speed and the planned assumed speed exceeds a preset speed threshold;

[0076] Periodic trigger condition: The time interval since the last planning reaches a preset time threshold;

[0077] Rule change triggering conditions: When changes affecting management data are detected in the building management system, such as changes to the current route;

[0078] Predictive triggering conditions: Based on dynamic environmental data / rule management data prediction models (machine learning models, event-driven models), predict environmental state changes that will affect the current path within a specific time period in the future. For example, based on historical dynamic environmental data, machine learning models can be used to predict future population density in the area, and based on building activity arrangements, event-driven models can be used to predict future path congestion.

[0079] Different priority weights are assigned to different triggering conditions. When multiple triggering conditions are met at the same time, they are processed in order of priority. For example, rule change trigger (highest) > user serious deviation trigger > environmental state change trigger > predictive trigger > periodic trigger (lowest). When multiple conditions are triggered at the same time, the highest priority trigger is processed according to priority.

[0080] The design incorporates a multi-dimensional intelligent triggering mechanism, including environmental mutations, user deviations, and periodic updates, coupled with replanning technology. This enables the system to continuously monitor during navigation and quickly and smoothly adjust the route once it detects that the path is no longer optimal. This forms a closed loop of "planning-execution-monitoring-adjustment," creating system robustness from "open-loop execution" to "closed-loop adaptation." This ensures the continuous optimality and robustness of the navigation service throughout the entire task cycle, resulting in a continuous and stable user experience.

[0081] Establish a closed-loop system of "planning-execution-monitoring-adjustment" and ensure the continuous optimality and reliability of the navigation path throughout the entire execution process through multi-dimensional intelligent triggering mechanisms (environmental changes, user deviations, periodic updates, etc.) and update and replanning technology. This effectively solves existing technical problems such as static planning becoming invalid once generated and the inability to correct errors in a timely manner when the environment changes or the user goes astray, leading to interruptions in the navigation experience.

[0082] In summary, this invention aims to solve the problems of inflexible cross-layer planning, single constraints, and lack of dynamic response in current indoor navigation technologies.

[0083] The method includes: receiving user requests containing start point, destination, and route selection preferences; acquiring and integrating static geographic data, dynamic status data, user context data, and building rule management data to generate unified spatial environment status data; constructing a real-time constraint set containing multi-dimensional constraints based on this; building an initial multi-layer topology map based on static geographic data, and using an edge weight update model to quantize the real-time constraint set into dynamic weights of each edge in the map to obtain a dynamically weighted multi-layer topology map; further, using a path evaluation model, comprehensively evaluating and filtering candidate paths based on the real-time constraint set, and outputting the optimal dynamic cross-layer planning path that satisfies multiple constraints; during navigation, monitoring the environment and user status through a multi-dimensional triggering mechanism to trigger dynamic replanning to maintain the path as continuously optimal.

[0084] It enables personalized path planning that integrates multiple constraints, performs real-time dynamic adjustments, and makes intelligent decisions across three layers in complex indoor environments, significantly improving the accuracy, robustness, and user experience of navigation.

[0085] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto; any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A dynamic cross-layer path planning method based on multiple constraints, characterized in that: Includes the following steps: S1. Receive user route planning request: Receive the user's input of the starting point, destination, and route selection preferences; S2. Cross-layer data acquisition and fusion: Acquire multi-source data of the target space, including static geographic data, dynamic environmental data, user context data and building rule management data, and align and fuse the above four types of data in time and space to form unified spatial environment status data; S3. Construction of multi-layer topology graph: S31. Based on static geographic data, the indoor multi-level space is abstracted into an initial multi-level topology graph containing nodes and connecting edges of each floor, and each connecting edge is assigned an initial weight. S32. Based on path selection preferences and spatial environment status data, extract and construct a real-time constraint set containing multiple quantitative constraints. S33. The constraints that can be assigned to the connecting edges in the real-time constraint set are quantified by a preset edge weight update model, and the comprehensive dynamic weight of each edge in the initial multi-layer topology graph is calculated and updated to obtain a dynamically weighted multi-layer topology graph. S4. Dynamic cross-layer multi-path generation: S41. On a dynamically weighted multi-layer topology graph, run a path search algorithm to generate multiple candidate paths from the starting point to the ending point. S42. Based on the real-time constraint set, multiple candidate paths are evaluated and screened as a whole through a pre-set path evaluation model, and the optimal dynamic cross-layer planning path that meets multiple constraints is output.

2. The dynamic cross-layer path planning method based on multiple constraints according to claim 1, characterized in that: In S2, static geographic data includes BIM or interior maps, including floor plans, room / corridor outlines, and locations of fixtures (stairs, elevator shafts).

3. Dynamic environmental data includes data from sensor networks, such as the real-time operating status of the elevator, passenger density inside the car, passenger density in the waiting area, and passenger density in key areas (corridors, escalator entrances and exits).

4. User context data includes the user's real-time location coordinates, movement speed, and historical route selection preferences.

5. Rule management data includes information from the building management system, such as temporary closure notices for areas, access permission lists for specific access control systems, and route control information resulting from large-scale events.

6. The dynamic cross-layer path planning method based on multiple constraints according to claim 2, characterized in that: In S31, key nodes include cross-floor facility nodes, which represent path decision points, such as key locations like corridor intersections, room entrances and exits, stairwells, and elevator lobbies. Edges include horizontal movement edges and cross-floor virtual edges connecting corresponding cross-floor facility nodes on different floors, representing passable paths between key nodes.

7. The dynamic cross-layer path planning method based on multiple constraints according to claim 3, characterized in that: In S32, based on path selection preferences and spatial environment status data, multi-dimensional constraints are extracted using a constraint building engine or constraint extraction logic to form a real-time constraint set. The real-time constraint set includes at least two of the following types of constraints: Edge-determinable constraints: Constraints that can be quantified and assigned to a single edge, including distance constraints based on the physical length of the path, time constraints based on real-time traffic conditions, comfort constraints based on the density of people in the area, and rule constraints set by the building management. Path-level constraints: Constraints that need to be statistically analyzed or evaluated at the complete path level, including waiting and transfer constraints based on the real-time status of cross-level facilities, path type consistency constraints based on user preferences, and total path energy consumption constraints.

8. The dynamic cross-layer path planning method based on multiple constraints according to claim 4, characterized in that: In step S33, the edge weight update model calculates a comprehensive dynamic weight value for each connecting edge based on the edgeizability constraint; in step S42, the path evaluation model generates an evaluation dimension score for each candidate path based on the edgeizability constraint and the path-level constraint, and selects the candidate path with the highest score as the dynamic cross-layer planning path.

9. The dynamic cross-layer path planning method based on multiple constraints according to claim 5, characterized in that: Also includes: S5, Path Dynamic Replanning: Outputs a dynamically planned cross-layer path and continuously monitors it during navigation, updates the real-time constraint set, performs dynamic replanning based on a preset real-time constraint triggering mechanism, determines whether the preset triggering conditions are met, and if so, re-executes S32-S42 to obtain the current optimal navigation path.

10. The dynamic cross-layer path planning method based on multiple constraints according to claim 6, characterized in that: The real-time constraint triggering mechanism preset in S5 includes, but is not limited to, the following triggering conditions: Environmental state change trigger condition: The change value of dynamic environmental state parameters related to the current navigation path exceeds the preset parameter threshold; User deviation trigger conditions: The deviation distance between the user's actual location and the planned path exceeds a preset distance threshold, or the difference between the user's actual travel speed and the planned assumed speed exceeds a preset speed threshold; Periodic trigger condition: The time interval since the last planning reaches a preset time threshold; Rule change triggering conditions: When changes affecting management data are detected in the building management system, such as changes to the current route; Predictive triggering conditions: Based on dynamic environmental data / rule management data prediction models (machine learning models, event-driven models), predict environmental state changes that will affect the current path within a specific time period in the future.

11. The dynamic cross-layer path planning method based on multiple constraints according to claim 7, characterized in that: Different priority weights are assigned to different triggering conditions. When multiple triggering conditions are met at the same time, they are processed in order of priority.

Images