Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
  • G01C21/20—Instruments for performing navigational calculations
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
  • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
  • G05D1/0005—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot with arrangements to save energy
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
  • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
  • G05D1/10—Simultaneous control of position or course in three dimensions
  • G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
  • G05D1/106—Change initiated in response to external conditions, e.g. avoidance of elevated terrain or of no-fly zones
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
  • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
  • G05D1/10—Simultaneous control of position or course in three dimensions
  • G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
  • G05D1/106—Change initiated in response to external conditions, e.g. avoidance of elevated terrain or of no-fly zones
  • G05D1/1062—Change initiated in response to external conditions, e.g. avoidance of elevated terrain or of no-fly zones specially adapted for avoiding bad weather conditions
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
  • G06Q10/00—Administration; Management
  • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
  • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
• • G—PHYSICS
  • G08—SIGNALLING
  • G08G—TRAFFIC CONTROL SYSTEMS
  • G08G5/00—Traffic control systems for aircraft, e.g. air-traffic control [ATC]
  • G08G5/003—Flight plan management
  • G08G5/0034—Assembly of a flight plan

Definitions

• Disclosed is a method and system for optimizing the path of a vehicle in multi -dimensional space.
• aircraft flight path optimization is disclosed.
• Predetermined air routes are often aligned with ground-based navigational aids. In some cases, air routes circumvent geographical regions. Great circle routes, on the other hand, promise shorter flight distances.
• Aircraft efficiency improves with favorable winds.
• ground speed increases and fuel consumption drops.
• Reduced fuel consumption often means that additional revenue-generating payload can be carried.
• Increased ground speed means that flight times are reduced resulting in operational cost savings.
• hazardous weather can impose a wide variety of costs on aircraft operations. Such costs can range from an uncomfortable ride for passengers at the low end, to structural damage, and even loss of aircraft and lives, at the other extreme. Aircraft operators typically go to great length to avoid hazardous weather.
• Achieving a desired arrival time is important because it allows the operator to more accurately schedule flights and enjoy greater operational efficiency. Aircraft operating on pre-determined air routes or great circle routes may be forced to make costly adjustments to airspeed in order to meet scheduling requirements.
• Typical flight path routing fails to consider operational costs associated with atmospheric phenomena, hazardous weather avoidance, region avoidance and arrival time constraints. There exists a need for a system that addresses these shortcomings.
• the invention determines an optimized route between a starting location and a destination by minimizing a cost function.
• the cost function is based on factors including fuel consumed, time, over-flight fees and penalty fees associated with restricted airspace.
• the invention implements an optimization algorithm for determining both lateral and vertical planes.
• the routes may be displayed over a global map with data including overlays depicting wind fields and restricted regions. Global databases may provide the data for wind conditions, weather, and temperature profiles.
• the invention determines the lateral path using a dynamic programming algorithm.
• the algorithm operates by constructing a rectangular grid that overlays both the origin and destination location. Determination of the optimal lateral path proceeds in stages beginning with the initial, or origin location. At each stage, the algorithm examines the transition cost from the current node to each of a set of possible nodes.
• the vertical path is determined using calculus of variations methods. Calculus of variations methods is a generalized optimization technique and is known in the art. In one embodiment, the vertical path is based on predetermined data stored as a function of parameters including wind speed, aircraft weight and cost index.
• a filter is used to smooth the route while it is being generated. Filtering improves the quality of the route by minimizing the effects of quantization. Quantization of the calculation may result in an optimal route that requires excessive heading and altitude changes.
• the user can select from among various routing choices. For example, wind-optimization routing results in substantially maximizing the favorable effects of taihvinds and substantially minimizing the detrimental effects of headwinds.
• Region avoidance routing provides an optimal route that circumvents specified geographical regions.
• over-flight routing provides an optimal route that considers over- flight costs in calculating a route.
• Required time of arrival routing provides an optimal route that achieves a desired arrival time.
• Figure 1 is an example of a routing scenario.
• Figure 2 is a global view illustrating a rectangular state space search region.
• Figure 3 is a map showing wind contours, a great circle route and a wind optimal route.
• Figure 4 is a map showing exemplary routes circumventing various restricted regions.
• Figure 5 is a grid illustrating path development between nodes A and B.
• Figure 6 is a grid illustrating possible departure segments from selected nodes.
• Figure 7 is a larger scale view of a portion of Figure 2.
• Figure 8 is a global view depicting wind fields over 50 knots at a particular altitude.
• Figure 9 is a larger scale view of a portion of Figure 8.
• Figure 10 is a global view illustrating a great circle route and a wind- optimized route between a node at an origin location and a node at a destination location.
• Figure 11 is a global view illustrating exemplary routes between a node at an origin location and a node at a destination location, in light of an avoidance region.
• Figure 12 is a view illustrating exemplary routes between a node at an origin location and a node at a destination location based on considerations of overfly costs and minimum fuel.
• Figure 13 is a global view illustrating a minimum fuel route and a minimum fuel to satisfy a required arrival time route between a node at an origin location and a node at a destination location.
• Figure 14 is another view of a grid illustrating path development between nodes A and B.
• Figure 15 is a coordinate system illustrating one geometry for route planning.
• Figure 16 is an illustration of geometry with a state space grid superimposed.
• Figure 17A is an illustration of one embodiment of a segment numbering system.
• Figure 17B is an illustration of departure segments, corresponding to Figure 17A, based on an approach segment.
• Figure 17C is a tabular representation of the data of Figure 17 A.
• Figure 17D is a tabular representation of step increments as a function of departure segments.
• Figure 17E is a tabular representation of reciprocal segments.
• Figure 18 is an illustration of lengthened paths around a great circle.
• Figure 19 is a depiction of the vertical flight path regimes.
• Figure 20 is an illustration of quantization effects.
• Figure 21 is an illustration of departure segments at a node.
• Figure 22 is an illustration of path filtering.
• Figure 23 is an illustration of the effect of filtering on route selection.
• Figure 24 is a block diagram illustrating implementation of one embodiment of the invention.
• a node represents a physical position in space corresponding to latitude, longitude and altitude (or other such coordinates).
• location is reserved generally for the node associated with the origin or destination and corresponds to the beginning and ending, respectively, of a particular route.
• the description is divided into multiple sections.
• the first section provides an overview and an example of route optimization.
• the second section describes one embodiment of lateral path optimization and application to flight path optimization and includes information showing possible results.
• the third section describes one embodiment of vertical path optimization as applied to various flight regimes.
• the fourth section describes one embodiment of path filtering and quantization.
• the fifth section describes an embodiment of a computer system adapted for implementing the invention.
• the invention relates to determining optimal path routing.
• the horizontal, or lateral, path is optimized by an iterative process of evaluating a plurality of alternative routes.
• the alternative routes are defined by the interconnections between nodes.
• the position of the nodes is arbitrary.
• the nodes are positioned at predetermined points arranged in geometric relation to the position of the origin and destination locations.
• the nodes may be distributed uniformly and arranged in a geometric grid pattern.
• the position of subsequent nodes may be defined in relation to the position of previous nodes.
• the nodes may be positioned, or defined, concurrent with the development of an optimized route, thereby, eliminating the requirement of a grid.
• the nodes may be arranged in concentric segments of semi-circles.
• the first set of nodes may consist of the set of points lying on a one mile radius spaced at 10 degree arcs, along a 180 degree semi-circle.
• subsequent sets of nodes may be defined at subsequent arcs projecting radially from the nodes lying on the optimized route.
• Figure 1 depicts an example of traveling from location A to location B.
• Grid 100 encompasses the space between the two locations and is of rectangular configuration. For this example, travel can proceed along any one of three possible paths where each path is made up of segments.
• segment refers to a portion of the flight path between nodes and the term “segment” may be further modified as in “approach segment” and "departure segment” wherein the directional term is with respect to a particular node under consideration.
• the available routes are segments 110, 130, and 150, leading to nodes marked x, y, and z respectively.
• stage refers to the transition from one node to a group of nodes.
• the path options can be denoted as A-x-B, A-y-B and A-z-B.
• a transit cost can be calculated for arriving at node x, node y, and node z. Similarly, a cost can be determined for each possible travel path that finishes at location B. Assume that arrival at node x carries a cost of $1.00, y is $2.00 and z is $3.00 and arrival at location B adds another $1.00. Consequently, path A-x-B will cost $2.00, path A-y-B is $3.00, and path A-z-B is $4.00. It will be seen that path A-x-B carries the least cost.
• node x may be enjoying favorable tailwinds
• node y may be experiencing a turbulent weather
• node z may be associated with burdensome over-fly costs.
• the simple grid of Figure 1 is replaced by a larger grid having many nodes.
• the grid is sometimes referred to as a state space region, an example of which is seen in Figure 2.
• state space region 100 is overlaid atop global map 5. Flight is contemplated between location A and location B.
• Route optimizing consists of identifying the path that yields the desired results. Desired results might include, but are not limited to, a route that requires the least fuel, least travel time, and the avoidance of geographical and political regions. In some situations it may be desirable to select a route that satisfies a required time of arrival. In addition, desired results might consist of various combinations of these, or other, objectives.
• Figure 3 illustrates the effects of wind optimization for a domestic flight across the United States.
• the figure depicts a generally eastbound flight between location A (Los Angeles ) and location B (Washington D.C.).
• Wind contour lines 10 also generally eastbound, overlay the US map 5 and depict the wind velocity at a particular flight level.
• Great circle route 30 and wind optimal route 40 are marked on map 5.
• the difference in flight path routing demonstrates the effects of wind optimization.
• Figure 4 illustrates the effects of a second optimization option, that is, region avoidance optimization.
• the figure depicts a generally eastbound flight across the Atlantic Ocean between location A, a Texas location and a location B, a European location. Between the origin and destination location are a number of regions that are advantageously avoided. Special Use airspace 20, severe weather regions 35 and 45 and political region 55 present obstacles to a typical flight. Region avoidance routing capabilities of the invention allows the user to select an optimized route that circumvents these regions. Route 60 and route 70 depict possible solutions as generated by this function. In practice, the user will provide further criteria to enable the invention to make a selection between these alternatives. Such further criteria may include wind contour lines or a required time of arrival. Various types of regions may be avoided using this function, including environmentally sensitive regions. In addition, the invention can accommodate changes in position, area, and degree of danger or cost associated with the avoidance region.
• Route optimization considers multiple paths throughout the grid where each path is defined by the segments between nodes. Consistent with the mode of optimization, the algorithm selects the least cost route between the origin and the destination as connected by the segments.
• FIG. 1 provides an illustration showing the segmented route generated by this method for travel between location A and location B.
• Grid 100 provides a reference for the plurality of nodes of which node x, node y, and node z are marked.
• Segments connecting various nodes are marked 110, 120, 130, 140, 150, and 160.
• the example treated node A as having three possible departure segments and nodes x, y, and z each having a single departure segment.
• each node has nine possible departure segments for every approach segment.
• Figure 6 illustrates one variety of possible departure segments, wherein node A is the origin location from which nine departure segments are possible.
• Node x and node y can be reached from node A in a single segment.
• this embodiment of the invention provides nine possible departure segments from that node.
• from node y there are nine possible departure segments available from node k based on the approach segment to node k (denoted as segment 120).
• Air travel also includes a vertical path and this too can be optimized.
• Parameters associated with the optimization of the vertical path include wind speed, aircraft weight, and cost index.
• the cost index serves as a model of the aircraft performance. Based on these parameters (including wind speed, aircraft weight, and cost index) there is a corresponding optimum cruise altitude, speed and fuel flow rate.
• the vertical path is optimized using a stored look-up table.
• optimization necessitates a balancing of these factors. For example, optimization based solely on wind data may provide a route that carries substantial over-flight fees, violates restricted airspace and may actually reach the destination too late. A different optimization may strike a satisfactory balance in that wind advantages are substantially maximized and disadvantages are substantially minimized, commensurate with substantially minimized over-flight fees and avoidance of restricted airspace (regions) in which the flight satisfies a required time of arrival.
• routes can be displayed over a world map as in Figure 3. Overlays may include wind fields, restricted regions and regions associated with over-flight fees as in Figure 4. Fuel requirements as well as en route times can be displayed and compared for various routes.
• Comparisons can be made with great circle and other routes.
• typical modes of operation for the invention include wind optimization, over-flight fee optimization, region avoidance, and meeting a desired time of arrival.
• the user specifies the origin and destination locations, the parameters (or parameters) to be optimized, and selects a viewing option.
• Users of the invention may include the aircraft pilot or other flight planning personnel. Viewing options might include selecting from among various depictions of the globe or a map and may include options for image rotation and zooming.
• the user also may be presented with various options for saving calculated routes. Other options allow for retrieval and display of previously saved routes.
• These functions allow the user to compare various routes. Other functions may allow the user to display multiple routes simultaneously or overlay such data as wind maps or restricted regions. In one embodiment, an assortment of features allows the user to compare routes based on forecasted, as well as actual, weather.
• the invention utilizes a dynamic programming method for determining an optimal lateral route.
• the following description provides an overview and additional detail regarding the dynamic programming method. More detailed descriptions of dynamic programming appear in later sections of this application.
• a state space search region (or "grid") is generated encompassing these locations.
• Figure 2 shows a state space search region for a flight from Minneapolis to Paris.
• Figure 7 depicts a smaller scale view of a portion of this state space region 100.
• Figure 7 illustrates one embodiment of a node arrangement, wherein node x, node y and node z are marked atop a portion of map 5.
• the grid encompasses the region to be searched for the optimal flight path and has dimensions N ⁇ by N ⁇ .
• the size of the grid is a balance between the computation time and range of options for a suitably optimized route.
• the invention will specify a route that follows along a boundary of the grid. This indicates that the boundary of the grid is serving as an artificial limitation on the optimization of the route.
• the grid is too large, then computation time is longer.
• the grid extends a short distance beyond the origin A and destination B locations to allow for advantageous, but circuitous routes.
• the grid includes a large number of latitude and longitudinal lines and has a node at each intersection. The invention optimizes a route of inter-connected nodes throughout the grid.
• the method implemented by one embodiment of the invention is as follows: 1. Beginning with the origin location, an approach segment is assumed and nine departure segments are generated. This is the first stage. For each approach segment there are nine possible departure segments. The incremental cost for each transition is calculated. The nodes thus generated become active nodes. Active nodes are nodes eligible for further extension and consideration as part of the solution to the optimization process. 2. At each active node generated by the first stage, a set of nine departure segments are generated. This is the second stage. The incremental cost for each transition is again calculated. The total cost to the next node is calculated by adding the incremental cost to the accumulated cost. The recursive equation for the accumulated cost is given by:
• the process continues in stages. For each active node, a set of nine possible departure segments are generated. The node which was extended then becomes inactive. An inactive node can become active again if reached by an extension and the new cost is lower than the previous cost to reach that node. This is implemented by providing an even stage active node list and an odd stage active node list. This technique allows one active node list to be retained while the other is being generated.
• the cost is compared to the cost at the destination node. If the cost at the node is larger than the cost to the destination node, then the node is removed from the active node list. Initially, the destination node is assigned a sufficiently high cost. If the assigned cost to arrive at the destination node is too low, then no other path will be found to be lower. In this case, the optimized cost to arrive at the destination node will match the assigned cost to arrive at the destination node. To prevent such a result, optimization should be repeated using a higher cost to arrive at the destination node. The process continues until no active nodes remain.
• the path is then generated by starting at the destination node and proceeding backwards from the destination location to the origin location using the stored direction of approach associated with each node in the path.
• the invention may incorporate data from a variety of sources in determining the cost to arrive at any particular node.
• the invention may utilize global databases showing regions having over-flight fees and restricted airspaces.
• the invention may also utilize data showing wind and atmospheric information.
• FIG. 8 depicts a global view of prevailing winds exceeding 50 knots at a selected altitude.
• Wind information is represented by wind barbs.
• Figure 9 depicts a closer view of such a wind barb 165 in a particular geographic region.
• Each wind barb 165 denotes wind direction and magnitude.
• the magnitude is indicated by the tail 170 on the barb, wherein a triangle 175 indicates 50 knots and each line 180 indicates an additional 10 knots.
• Wind direction is denoted by the barb direction indicated as 185.
• optimization choices available to the user include the following:
• Figure 3 illustrates the effects of wind optimization on routing.
• Figure 10 provides map 5 illustrating another example of wind optimal routing.
• origin location A represents Los Angeles and destination location B represents London.
• Route 240 is the shortest distance corresponding to the great circle route.
• Route 250 is a wind-optimized route that requires less fuel and less en route time.
• Region Avoidance Routing illustrates the effects of region avoidance optimization.
• Figure 11 provides another example of region avoidance routing. The illustration depicts a flight from location A in Minnesota to location B in Paris wherein a hazardous weather system lies along the path between the two cities. Flight route 260 circumvents the weather system with minimal fuel and flight route 270 passes through the weather system 275 with minimal fuel.
• Over-Flight Fees A third optimization can be performed to minimize over- flight fees. Certain political regions levy over-flight fees for transient aircraft. Over-flight fees may be an attempt to mitigate harm from noise and pollution. Routing of aircraft to avoid such fees can be cost effective. The subject matter calculates routes based, in part, on the over-flight fees. Routes minimizing over-flight fees can be compared with alternative routes offering such advantages as reduced transit time or reduced fuel consumption.
• Figure 12 illustrates an example of a route configured with regard to over-flight fees.
• a minimum fuel route from location A (Anchorage) to location B (New York) with over-flight fees over Canada 220 is shown.
• An over-flight fee is assumed to be charged for flight in Canadian airspace.
• Three routes are shown: (a) a minimum fuel, labeled 190; (b) minimum fuel with Canadian overflight fees 210; and (c) minimum fuel with a specified time of arrival 200.
• a fourth optimization function enables the user to specify a required arrival time. Accurate flight scheduling reduces operating costs and therefore, efficiency.
• the invention enables route optimization that capitalizes on the favorable effects of tailwinds while avoiding unfavorable headwinds, circumvents weather or other restricted airspace, and yet achieves the required time of arrival.
• Figure 13 depicts the effect of arrival time on aircraft- routing.
• Route 230 provides the lowest fuel requirement but at the cost of a delayed arrival.
• Route 220 satisfies the arrival time requirement at a minimum fuel consumption.
• Dynamic routing enables determination of a new route based on the present aircraft position and the desired destination. Such in-flight dynamic routing can be performed by ground-based operators or airborne personnel. As with the aforementioned modes of optimization, dynamic routing may include altering the lateral path, the vertical path, and speed profiles.
• J k+] min ⁇ L[g(x(k - l)),u(k),k)] + J k ⁇ u(K)
• Figure 14 depicts the dynamic programming approach.
• Figure 14 represents routing expressed in coordinates x, and x 2 where the origin location A is denoted as x(l) and the destination location B is denoted as x(N).
• Path 300 and 310 are alternative routes approaching the destination location B.
• the recursive optimization function being implemented is
• J k+1 min ⁇ L[g(x(k - l)), u(k), k)] + J k ⁇ u( )
• the previous stages are represented as x (1) (k) having cost J ⁇ j . and x (2) (k) having cost J 2* ⁇ .
• the decisions and transition costs can be represented as ⁇ C[x(k), u(k),k].
• a stage is defined as the step to the next set of nodes.
• the optimal control u(k), and the corresponding performance value J k is derived from the recursive optimization function. If the state has been reached before retain the lowest accumulated cost at the node and the corresponding approach segment. 5) The process ends when the desired destination state has been reached and the cost to other states is greater. Determine the total accumulated cost to the state from the recursive optimization function. 6) The optimal path corresponding to the initial state x(l) is recovered by inserting the optimal control policy. Dynamic programming is advantageous because it finds a global solution and accounts for state variable constraints. Part C - Application to flight path optimization
• N grid N v x N ⁇ x N ⁇ x N h x N ⁇ x N ⁇ x N mf .
• Such a grid would include an extremely large number of nodes. Practical approximations can reduce the dimensions of the grid and yet render a reasonable solution with a marginal effect on accuracy.
• the integral cost function can be expressed in a recursive form as follows:
• the lateral acceleration a ⁇ is selected to make the transition to the new latitude-longitude position.
• the remaining control variables are chosen to optimize the transition costs ( ⁇ C).
• the route solver performs the optimization calculations.
• one embodiment of the route planning geometry is shown in Figure 15.
• Figure 15 depicts a three-axis coordinate system of X e by Y e by Z e .
• Angle ⁇ and angle ⁇ denote the longitude and latitude, respectfully, and N denotes the velocity.
• Other variables in the various equations are assigned the following definitions:
• the objective of minimizing, or achieving, a desired parameter is expressed in the form of a cost function.
• the cost function may contain a number of terms, including fuel, time, weather risk, over-flight fees and restricted areas.
• Variables denoted as K are gain factors and Cl denotes the fuel- time cost index.
• the cost index is a function of the aircraft and engine performance and operating environment.
• V — gsin ⁇ + ⁇ c Rcos ⁇ (sin ⁇ cos ⁇ - cos ⁇ sin ⁇ sin ⁇ ) m
• Vehicle and atmospheric models are utilized to solve the aforementioned equations. These models can be expressed as follows: Vehicle Models
• T e T e ( ⁇ , ⁇ ,h) temperature profiles
• V w V w ( ⁇ , ⁇ ,h) wind profiles
• P P(T e ) density
• the parameter ⁇ is limited by the maximum allowable acceleration of the aircraft.
• FIG. 16 The geometry for the route solver algorithms is shown in Figure 16. Much like the geometry of Figure 15, a three-axis geometry 50 is again depicted. In Figure 16, the origin location A has spherical coordinates longitude ⁇ 0 and latitude ⁇ 0 shown in geometry 50, and the destination location B has spherical coordinates longitude ⁇ d and latitude ⁇ d shown in geometry 50. In one embodiment of the invention, these standard earth centered spherical coordinates are then transformed to the rectangular geometry denoted in grid 100 of Figure 16 wherein origin A' has coordinates i 0 and j 0 (cooresponding to origin A) and destination B' has coordinates i d and j d (corresponding to destination B).
• the rectangular geometry of grid 100 of Figure 16 may be thought of as a spherical coordinate system with the equator positioned between the origin and destination.
• the weather profiles and avoidance region coordinates are also transformed to rectangular geometry. This transformation makes the grid more dimensionally uniform.
• the route coordinates are then transformed back to the standard earth centered spherical coordinate frame.
• the coordinates are made discrete according to:
• the steps performed in the route solver algorithms are as follows: 1) In one embodiment, a pair of active node lists are generated. An even active node list and an odd active node list are generated so that the old active node list can be retained while the new active node list is being generated.
• the active node list contains the location of the node as shown in the Table 1. At the origin location (first active node), nine heading directions are generated. This is the first stage. The incremental cost for each transition is calculated.
• Figure 17A graphically illustrates twenty-four radial directions and their corresponding numerical labels.
• Figure 17B illustrates the nine departure segments that are generated by the invention based on an entry direction corresponding to a segment marked 3.
• the segments depicted in Figure 17B correspond to the radials in Figure 17 A and are marked 1 , 5, 6, 11, 12, 19, 20, 4 and 2.
• the table of Figure 17C indicates the approach segments as "i" and the nine departure segments as "j".
• approach segment 3 corresponds to departure segments marked 1, 5, 6, 11, 12, 19, 20, 4 and 2. 29
• Figure 17D is a tabular representation of the segment corresponding to each numbered departure segment.
• departure segment 1 corresponds to segment 1 as shown in Figure 17A as well as segment 1 as shown in Figure 17B.
• Figure 17D provides that ⁇ x(i) is +1 and ⁇ y(i) is 0.
• the ⁇ x(i) and ⁇ y(i) steps for each of twenty- four exit directions are shown in Figure 17D.
• step refers to a change of position relative to the node under consideration and, in this embodiment, is described in rectilinear coordinates.
• segment 13 corresponds to a segment that lies between the node under consideration and the node located relative to +2 in the
• Figure 17E is a tabular representation of reciprocal radials for each segment.
• departure segment 1 corresponds to approach segment 3 and the reciprocal of segment 10 is segment 14. After extending a node it becomes inactive.
• the cost at each node reached (Cp ip p ) is compared to two other costs.
• the two other costs to be compared with Cp j are, first, the cost previously stored at that node, and second, the cost stored at the destination node. IfCp ip is less than both the cost previously stored at that node and the cost stored at the destination node, then the cost previously stored at that node is replaced with Cp ⁇ : p and the direction of approach to that node is also replaced with the direction of approach associated with Cp ip : p .
• Cp ip : is more than either the cost previously stored at that node or the cost stored at the destination node, then the cost, and the direction of approach, previously stored at that node are retained and Cp ip : p , and the corresponding direction of approach to that node, is discarded.
• the approach segment is determined by the direction of the departure segment from the previous node using the mapping table as shown in Figure 17C. If the cost at the node is larger than the cost to the destination node, then the node is removed from the active node list. 2) The process continues in stages. Inactive nodes can become active again if it is reached by an extension and the new cost is lower than the previous cost to reach that same node. The process continues until no active nodes remain. 3) The path is generated by starting at the destination node and proceeding backward to the origin location using the direction of approach for each node.
• a penalty cost is associated with nodes within the avoidance region.
• the cost function with the avoidance region cost included, is:
• the region is to be avoided at all costs, then a very large number is assigned to R( ⁇ , ⁇ ) within the region. If the penalty in the region is varying, then varying values can be assigned to the function R( ⁇ , ⁇ ) within the region. The factor K- is adjusted to change the proximity of the path to the higher penalty areas of the region.
• the nodes within a region to be avoided can be eliminated from the grid. By eliminating nodes in avoidance regions, the number of calculations required for optimization is reduced. Regions treated as avoidance regions include regions of severe weather, special use airspace, politically sensitive regions, and environmentally sensitive regions.
• the cost function including the different region types is:
• ⁇ C [ ] ⁇ S + K r R( ⁇ , ⁇ ) + K w W( ⁇ , ⁇ ) + K sua SUA( ⁇ , ⁇ ) + K D P( ⁇ , ⁇ ) + K e e( ⁇ , ⁇ )
• a required time is reached by adjusting the Cost Index (Cl).
• the route may be determined by iteration. First, an initial time cost index (Cl) is selected. The selection of an initial Cl can be made by the user or the computer may supply a value. Next, an optimized path is determined from which the time of arrival at the destination location is determined. The difference between the determined time of arrival at the destination location and the desired time is determined.
• the cost index is adjusted.
• the cost index carries a heavy weight on time.
• the cost index can only be reduced to a value corresponding to the maximum endurance cruise condition for that particular aircraft.
• the fuel flow rate is at a minimum and the optimum cruise airspeed is then a function only of aircraft weight.
• additional processes are used. Exemplary processes are as follows:
• the cost function in the lateral route solver is changed from minimum fuel and time to minimizing distance and time.
• the cost function increment is:
• An initial value of C s is chosen (by the user or the computer may supply a value) and an optimized route as well as time of arrival is determined.
• the time of arrival error (difference between the calculated time of arrival and the required time of arrival) is determined, and the factor C s is adjusted according to the time of arrival error:
• a set of paths around the great circle, shown in Figure 18, is defined by:
• the invention can provide advantageous routing as demonstrated by the following comparisons. It is noted, however, that these results are generated using assumed wind and atmospheric data as well as over-flight fees and avoidance regions representative of those existing at a particular time. As such, there exists no guarantee that future results will be similar. The following data shows results for these four cases:
• Wind Optimal Routes A wind optimal route for a commercial aircraft flying from location A (Los Angeles) to location B (London) is shown in Figure 10. The wind data is overlaid atop the image and is similar to the wind field illustrated in Figure 8. Great circle route 240 is also shown for comparison. The wind optimal route deviates to the south to take advantage of favorable tailwinds. A comparison of the fuel and time is given in' Table 2. Table 2: Wind Optimal Route Comparison of Fuel Used and Time
• the data shows a wind optimal route exhibits a savings of 6,145 pounds of fuel and 20 minutes of flight time.
• RTA Routes Wind optimal routes having a required time of arrival time (early and late) for a route between location A (Los Angeles) and location B (London) is shown in Figure 13.
• the wind data is overlaid atop the image and is similar to the wind field illustrated in Figure 8.
• the distance, fuel, and time is given in Table 3.
• the fuel and time for the RTA can be compared to the fuel and time for the no RTA case by referring to Table 2.
• the required times of arrival are met but with an increased cost of fuel.
• the required times are achieved by changes in both the lateral and the vertical profile.
• the average speed on the early time of arrival is increased over the minimum fuel route.
• the average ground speed on the late time of arrival is less that the minimum fuel route and there is a significant change in the lateral path.
• Region Avoidance The mimmum fuel route for a flight from location A (Minneapolis) to location B (Paris) with an avoidance region 275 shown in Figure 11.
• the cost associated with nodes in the avoidance region is set to a high value so that the region will be entirely avoided.
• Three routes are shown: (1) a minimum fuel route with no avoidance region - route 270; (2) a minimum fuel route with the avoidance region, route 260; and (3) a minimum fuel, required time of arrival route with the avoidance region, (similar to route 260).
• the distance, fuel and time are compared in Table 4.
• Over-Flight Fees A minimum fuel route for a flight from location A (Anchorage) to location B (New York) with over-flight fees in Canada is shown in Figure 12. An over-flight fee is assumed to be charged for Canadian airspace. Three route are shown: (a) a mimmum fuel, route 190; (b) mimmum fuel with over-flight fees, route 210; and (c) minimum fuel with a specified time of arrival, route 200.
• the minimum fuel route 190 is the great circle route.
• the mimmum fuel route with the over- flight fees 210 entirely avoids Canada.
• the middle route 200 is minimum fuel with over-flight fees and meeting a required arrival time. The distance, fuel and time is given in Table 5.
• the minimum fuel with fees (flight around Canada) route takes more fuel and time than the minimum fuel route but has the lowest total cost - even when over-flight fees are included.
• the minimum fuel with fees and RTA route takes less fuel and time than the "around Canada" route but has a higher cost including the fees.
• the transition cost ⁇ C is optimized.
• Optimizing of the transition costs ⁇ C implies optimization of the vertical path.
• optimization of the vertical path is performed independent of the optimization of the lateral path.
• the vertical path is divided into three regimes: 1) climb to subsomc cruise; 2) subsonic cruise; and 3) descent. These regimes are depicted in Figure 19. The following is a description of the assumptions and approximations used for determining the transition cost within, and between, each of these three regimes.
• ⁇ C cl ⁇ mb ⁇ m fchm b (W ⁇ o ) + CI ⁇ T chm b
• control variables are chosen to minimize this function and to satisfy the cruise equilibrium conditions of thrust equals drag and lift equals weight, i.e.,
• the aircraft performance model describing thrust, drag, and fuel flow rates are required for this computation. This optimization operation can be advantageously computed off-line.
• the fuel flow rate, cruise altitude and speed can be stored as a function of the parameters weight
• the aircraft weight, cost index and wind speed at the current aircraft location are used to determine the corresponding cruise altitude, speed, and fuel flow rate based on stored tables.
• the above approach shows a look-up table of three variables. A method of reducing this to a table look-up of two variables is given below.
• the weight and time at each node is determined from:
• the cruise solution is a function of three parameters: (a) wind speed; (b) Cost Index (Cl); and (c) aircraft weight (w).
• the cruise performance fuel flow rate, speed, and altitude
• the cruise performance may be stored as a function of these parameters, however, this requires storage of a large volume of data and three dimensional table look-up routines.
• a method to reduce the cruise performance cost from a three dimensional look-up to a two dimensional look-up is described. Following is a description of the method.
• the cost increment for the cruise to cruise transition is given by: ⁇ C + [ FFR+CI ] ⁇ S
• control variables are chosen to minimize this function and to satisfy the cruise equilibrium conditions of thrust equals drag, and lift equals weight but with the wind equal to zero as follows:
• This optimization operation is computed off-line and the fuel flow rate, the cruise altitude and speed are stored as a function of the parameters weight (W) and cost index (Cl).
• the tables that are stored include:
• V cn ⁇ V CTU (W,CI)
• FFR FFR(W,CI )
• V Supply V cm (W,CI )
• the state space quantization used in the dynamic programming method may lead to ambiguity in the solution, as illustrated in Figure 20.
• the cost function is the distance traveled between the nodes A and B.
• the permissible turn directions of the departure segment at each node is 0 , +45 and -45 relative to the direction of the approach segment. This means that at each node, the aircraft is allowed to turn leftward or rightward 45 or continue straight.
• the region of ambiguity is defined by the outer boundary of the permissible routes as noted by labels 400, 410, 420 and 430. Within this boundary there are a number of paths having equal lengths. A reduction of the grid quantization does not reduce the region of ambiguity. With a finer grid, the path from A' to B' exhibits the same characteristics as above with regard to path A to B.
• Figure 22 shows a grid 100 with an optimized path having three segments between node A and node B. The three segments are marked dO, dl, and d2.
• the filter would result in the selection of route 1 rather than route 2 as shown in Figure 23. This is a desirable result because post-filtering of the routes would result in a shorter distance using route 1.
• Filtering of the distance of the candidate routes results in a higher quality route.
• the filter minimizes the effect of the quantization in the lateral path.
• Figure 24 illustrates a block diagram of one embodiment of the invention 500.
• Route Solver 510 includes lateral path optimizer 560, vertical path optimizer 570 and time of arrival solver 580.
• Route solver 510 receives inputs from a variety of sources.
• Cost index 590 is a function supplied to route solver 510.
• Required time of arrival 600 is user-specified and also supplied to route solver 510.
• the user provides a site selection 670 that is coordinated with the city location database 610 which is supplied to route solver 510.
• Aircraft database 520 is supplied to route solver 510 and includes information about fuel flow rates, speeds and altitude data.
• Weather information 630 may be derived from a variety of weather sources 660 and includes temperature, wind, and storm cell information. The weather information 630 is supplied to the route solver 510.
• Data corresponding to restricted regions and costs 640 is supplied to the route solver 510.
• Data corresponding to over- flight fees 650 is also supplied to route solver 510.
• Route solver 510 generates the optimum route consistent with the user- specified parameters and conveys the solution to the display componentry.
• Monitor 540 is subject to control 550 and provides the means of displaying the route information.
• Monitor 540 derives data from the visualization processing module 530 which includes zoom, rotate and play-back features.
• the visualization processing module 530 receives data from route solver 510.
• the visualization processing module 530 receives also has available world map information 520.
• the invention may be adapted to operate on numerous computer platforms, including such systems as Silicon Graphics Indigo2 R4000/4400 work station or personal computers running such operating systems as Windows 95 or Windows NT.
• the user interfaces are functionally similar.
• the user interface includes Windows pull-down menus.
• Motif menus, check boxes, slide bars or any other such operator control functions can be utilized.
• Various display and control functions may be accessible via pull-down menus. Display capabilities of the work station or personal computer are adapted to depict multiple routes, land and water areas as well overlay selected weather information. Selected weather may include winds exceeding a specified speed.
• the disclosed subject matter includes encoding in a programming language such as FORTRAN or C. Processing elements available to the computer for determining optimal routing includes the following: 9
• This data may include a city location database or latitude and longitudinal coordinates. A user may input this data manually or make a selection from menus. Alternatively, a user may make a selection using a pointer and a map.
• This data provides a model of the aircraft operating performance.
• This data may include daily or hourly weather reports.
• Data may be measured or forecasted weather and includes global data bases depicting atmospheric, wind, weather, and temperature profiles as well as hazardous weather.
• Algorithms, routines, and data necessary for determining optimal vertical path routing based on selected parameters may be implemented as a look-up table or include arithmetic functions for optimizing the path.
• Cost Index model This model provides the operating parameters of the aircraft, or other vehicle. Aircraft modeling is a function of aircraft and engine performance data, including fuel flow rates.
• GUI graphical user interface
• the GUI includes modules that provides the visual appearance of the application and provide overall program control. These modules may be encoded in Motif-based C (for Silicon Graphics applications) or Microsoft Windows Visual C/C++ (for PC applications) or other such language.
• the graphics-hardware interface module may be written in C using Open GL graphics libraries to render three- dimensional displays. This module computes viewing perspectives, displays the globe, and draws path data on the display surface.
• the route-planning component computes mimmum time and minimum fuel routes between selected locations using optimization algorithms.
• Optimization may be performed on either the lateral plane, the vertical plane, or both planes.
• Suitable global databases characterizing such data as land and water areas, atmospheric, wind and temperature profiles are utilized by the route-planning component. In one embodiment, this function is written in FORTRAN and C.
• the user interface features Windows-type pull-down menus.
• Menu selections enable route generation and user-control of functions such as earth perspective, rotation and zoom.
• Other controls allow the user to save the current route or display previously-saved routes.
• Other maps including wind contour maps and land maps, may be overlaid atop the resulting routes via user-controlled check boxes and menus.
• Short-cut keyboard controls duplicate some of the pull-down menu functionality.
• the user is presented with various options for selecting origin and destination locations.
• Various options for identifying locations include, but are not limited to, pointing and clicking an icon on a map, or entering latitude and longitude coordinates, or selecting city names, or selecting from a database of locations.
• Such a database may consist of entries as shown below:
• This data may include daily or hourly weather reports.
• Data may be measured or forecasted weather and includes global data bases depicting atmospheric, wind, weather, and temperature profiles including hazardous weather.
• Wind data includes wind speed and direction as a function of latitude, longitude and altitude.
• Temperature profiles provide temperature data also, as a function of latitude, longitude and altitude. Such profiles are commercially available at six and twelve hour update intervals.
• One such service, the National Weather Service provides data in gridded binary format (GRIB) at 1.25° by 1J5 updated twice daily.
• Figure 8 depicts a global view of such wind data.

Priority Applications (3)

Applications Claiming Priority (2)

Publications (1)

Family

ID=22838190

Family Applications (1)

Country Status (5)

Cited By (17)

Families Citing this family (118)

Citations (4)

Family Cites Families (28)

• 1998
  • 1998-12-31 US US09/223,846 patent/US6266610B1/en not_active Expired - Lifetime
• 1999
  • 1999-11-16 EP EP99964993A patent/EP1141656A1/en not_active Withdrawn
  • 1999-11-16 CA CA002358477A patent/CA2358477A1/en not_active Abandoned
  • 1999-11-16 WO PCT/US1999/027150 patent/WO2000040929A1/en not_active Application Discontinuation
  • 1999-11-16 JP JP2000592599A patent/JP2004538438A/ja not_active Withdrawn

Patent Citations (4)

Also Published As

Similar Documents

Legal Events