Calculator overview
Inputs and outputs
This summary comes from the calculator's published input and output contract.
Inputs
- Time shares (row k is segment k; Manual shares only; a blank share counts as 1)
-
When omitted Blank
Default 0 rows
About this input
The time shares of the smooth planner with Manual shares: row k is segment k (in the order of the waypoints in use), each a number from 0.1 to 10 or blank (a blank share counts as 1). Segment k lasts T times its share over the sum of the shares in use.
Column Range or allowed values Time share 0.1 to 10 - Planner (required)
-
Default Smooth through the waypoints Allowed Trapezoid (stop at each), Minimum jerk (stop at each), Smooth through the waypoints
About this input
The planner: Trapezoid (stop at each), the time-optimal profile on each straight segment; Minimum jerk (stop at each), the rest-to-rest quintic at its shortest admissible time; or Smooth through the waypoints, a cubic in time per axis that passes every waypoint with continuous acceleration and starts and ends at rest.
- Waypoints (row i is waypoint row i; leave a row blank when it is not used) (required)
-
Default 4 rows
About this input
The waypoints: up to six rows of x and y in m, each from 0 to 4 m (the 4 m workspace, origin at the bottom-left corner). The rows in use are visited in row order and blank rows are skipped; at least two are needed, and consecutive waypoints must be at least 0.05 m apart (the first and the last may coincide).
Column Range or allowed values x (m) 0 to 4 y (m) 0 to 4 - Total time (smooth planner) (required)
-
Unit s Default 9 Range 1 to 60
About this input
T, the smooth planner's total time in s. The stop-at-each planners always run at their shortest times.
- Tracked mass (required)
-
Unit kg Default 2 Range 0.1 to 100
About this input
m, the mass of the body that tracks the plan, in kg.
- Timing rule (smooth planner) (required)
-
Default Chord length Allowed Uniform, Chord length, Manual shares
About this input
How the smooth planner shares its total time between the segments: Uniform, Chord length (in proportion to the segment lengths) or Manual shares (the time shares). Not used by the stop-at-each planners.
- Speed limit (required)
-
Unit m/s Default 1 Range 0.1 to 5
About this input
vmax, the speed limit in m/s.
- Challenge number (required)
-
Default 0 Range 0 to 5
About this input
Selects a challenge: 0 for none, 1 to 5 for a predefined challenge whose settings replace the inputs it fixes.
- Your prediction (challenge 1) (required)
-
Default Not yet predicted Allowed Not yet predicted, Trapezoid first, Minimum jerk first, Smooth through first
About this input
Your prediction for challenge 1: Not yet predicted, Trapezoid first, Minimum jerk first or Smooth through first. Not used outside challenge 1.
- Acceleration limit (required)
-
Unit m/s2 Default 1.5 Range 0.1 to 10
About this input
amax, the acceleration limit in m/s2.
- Force limit (required)
-
Unit N Default 20 Range 0.1 to 1000
About this input
F, the largest control force magnitude in N; a larger commanded force is scaled down to F, keeping its direction.
- Loop damping ratio (required)
-
Default 0.7 Range 0.2 to 2
About this input
zeta, the PD loop's damping ratio.
- Loop bandwidth (required)
-
Unit rad/s Default 4 Range 0.5 to 50
About this input
wc, the PD loop's natural frequency in rad/s: Kp = m wc^2 and Kd = 2 zeta m wc.
- Feedforward (required)
-
Default With feedforward Allowed With feedforward, Feedback only
About this input
With feedforward (the controller adds m times the planned acceleration to its PD correction) or Feedback only.
Outputs
- Peak acceleration
-
Unit m/s2
About this output
The selected planner's peak acceleration magnitude in m/s2 (exact).
- Segments of the selected plan: the waypoint rows joined, the start, duration and length, and the peak speed, peak acceleration and acceleration effort on each segment
-
No unit declared
About this output
The segments of the selected plan: one row per segment in use (5 rows; unused rows blank), with the waypoint rows it joins (for example 1 to 3), its start time, duration and length, and its peak speed, peak acceleration and acceleration effort.
- Motion frames: 201 frames k = 0 to 200 of the selected plan and the tracked body (the reference point, the body, the reference speed and acceleration, the tracking error, the applied force and 1 where the force limit acts)
-
No unit declared
About this output
The motion frames: 201 rows, one per frame k = 0 to 200 at equal steps of the selected plan, each with the time in s, the reference point x and y in m, the tracked body x and y in m, the reference speed in m/s and acceleration in m/s2, the tracking error in m, the applied force x and y in N, and 1 where the force limit acts (else 0).
- Peak speed
-
Unit m/s
About this output
The selected planner's peak speed in m/s (exact for the stop-at-each planners; sampled at 65 points per segment for the smooth planner).
- Peak jerk
-
Unit m/s3
About this output
The selected planner's peak jerk magnitude in m/s3; blank for the trapezoid, whose jerk is unbounded.
- Peak control force
-
Unit N
About this output
The largest applied control force magnitude over the 200 held steps, in N.
- Largest tracking error
-
Unit m
About this output
The largest distance between the body and the reference over the 201 frames, in m.
- Loop damping ratio used
-
No unit declared
About this output
The loop damping ratio used (the challenge's when a challenge fixes it).
- Loop bandwidth used
-
Unit rad/s
About this output
The loop bandwidth used, rad/s (the challenge's when a challenge fixes it).
- Formula module versions
-
No unit declared
About this output
The formula modules the model is built from, with their versions.
- Model version
-
No unit declared
About this output
The model version.
- Model Status
-
No unit declared
About this output
NOT VALID with the reason when an input cannot be used; CHECK when the results are shown but the sampled tracking loop is unstable, the bandwidth is high for the sample step, or a smooth segment lasts less than 0.1 s; otherwise OK. A plan over a limit, a force limit that acts and a path that leaves the workspace are results, not warnings.
- Planner used
-
No unit declared
About this output
The planner used (the challenge's when a challenge fixes it).
- Total time
-
Unit s
About this output
The selected planner's total time, in s.
- Timing rule used
-
No unit declared
About this output
The timing rule used (the challenge's when a challenge fixes it).
- Speed limit used
-
Unit m/s
About this output
The speed limit used, m/s (the challenge's when a challenge fixes it).
- Largest distance outside the workspace
-
Unit m
About this output
How far the selected path goes outside the workspace at most, in m (0 when inside).
- Path length through the waypoints
-
Unit m
About this output
The length of the straight path through the waypoints in use, in m.
- Tracked mass used
-
Unit kg
About this output
The tracked mass used, kg (the challenge's when a challenge fixes it).
- Scale factor to the limits
-
No unit declared
About this output
The uniform time-scale factor MAX(peak speed/vmax, SQRT(peak acceleration/amax)): stretching every duration by it makes the same plan just meet both limits. Blank while challenge 1 awaits a prediction.
- RMS tracking error
-
Unit m
About this output
The root mean square tracking error over the 201 frames, in m.
- The three planners on the same waypoints: total time, time at the limits, peak speed and acceleration, and the acceleration and jerk efforts
-
No unit declared
About this output
The three planners on the same waypoints: for each, its total time, its time at the limits, its peak speed and acceleration, and its acceleration and jerk efforts (the jerk effort blank for the trapezoid). Blank while challenge 1 awaits a prediction.
- Speed check bound
-
Unit m/s
About this output
The most by which the smooth planner's true peak speed can exceed the sampled peakSpeed, in m/s (its peak acceleration times the sample spacing over 2); 0 for the stop-at-each planners, whose speed is exact.
- Total time used
-
Unit s
About this output
The smooth planner's total time used, s (the challenge's when a challenge fixes it).
- Segments in use
-
No unit declared
About this output
The number of segments (waypoints in use less 1).
- Within the limits
-
No unit declared
About this output
Yes when the selected plan keeps both the speed and the acceleration limits (within 1E-9 relative), else No.
- Challenge goal
-
No unit declared
About this output
The goal of the selected challenge (blank with none).
- Challenge feedback
-
No unit declared
About this output
Feedback on the selected challenge: the prediction prompt, the met text or the not-met hint (blank with none).
- Path box, smallest y
-
Unit m
About this output
The smallest y the selected path reaches, in m.
- Challenge result
-
No unit declared
About this output
No challenge selected, MET or NOT MET.
- Challenge score
-
No unit declared
About this output
The challenge score: 100 when met, else 0.
- Challenge: the value scored
-
No unit declared
About this output
The value the selected challenge scores: the fastest planner's code 1 to 3 (1), the total time in s (2), the acceleration effort (3), the largest tracking error in m (4) or the distance outside the workspace in m (5); blank when not valid or withheld.
- Bandwidth times step
-
No unit declared
About this output
wc times h; above 0.5 the sampled controller no longer behaves like a continuous one.
- Acceleration limit used
-
Unit m/s2
About this output
The acceleration limit used, m/s2 (the challenge's when a challenge fixes it).
- Acceleration effort
-
Unit m2/s3
About this output
The acceleration effort, the integral of the squared acceleration magnitude over the plan, in m2/s3.
- Path box, smallest x
-
Unit m
About this output
The smallest x the selected path reaches, in m (exact: the end points and the turning points of each segment).
- Path box, largest y
-
Unit m
About this output
The largest y the selected path reaches, in m.
- Path box, largest x
-
Unit m
About this output
The largest x the selected path reaches, in m.
- Control energy
-
Unit N2 s
About this output
The control energy, the integral of the squared applied force over the plan, in N2 s.
- Path inside the workspace
-
No unit declared
About this output
Yes when the selected path stays inside the 4 m workspace (within 1E-9 m), else No.
- Share of steps at the force limit
-
No unit declared
About this output
The share of the 200 held steps in which the force limit acts, 0 to 1.
- Time at the limits
-
Unit s
About this output
The time at the limits: the total time multiplied by the scale factor, in s. Blank while challenge 1 awaits a prediction.
- Limit check
-
No unit declared
About this output
Within both limits, Over the speed limit, Over the acceleration limit or Over both limits.
- Jerk note
-
No unit declared
About this output
For the trapezoid: its jerk is unbounded, because it switches its acceleration instantly. Blank for the other planners.
- Jerk effort
-
Unit m2/s5
About this output
The jerk effort, the integral of the squared jerk magnitude over the plan, in m2/s5; blank for the trapezoid.
- Sampled loop stable
-
No unit declared
About this output
Yes when the sampled PD loop is stable at this step (zeta wc h < 1 and wc h < 4 zeta, the second-order Jury conditions), else No.
- Control sample step
-
Unit s
About this output
h, the tracking sample step: the total time over 200, in s.
- Force limit used
-
Unit N
About this output
The force limit used, N (the challenge's when a challenge fixes it).
- Tracking error at the end
-
Unit m
About this output
The tracking error at the last frame, in m.
- Feedforward used
-
No unit declared
About this output
The feedforward setting used (the challenge's when a challenge fixes it).
- Fastest planner at the limits
-
No unit declared
About this output
The planner with the least time at the limits on these waypoints (ties within 1E-9 relative go to the lower list position). Blank while challenge 1 awaits a prediction.
LogicCommons is in beta. If a result, label, or reference looks wrong, tell us here; we read every message.
Methodology
Purpose and model boundary
This lab plans the motion of a point body through 2 to 6 waypoints on a flat square floor 4 m across, and then has a body of a given mass follow that plan under a feedback controller. Three motion plans are set side by side on the same waypoints: a trapezoid that stops at each waypoint, a minimum-jerk law that stops at each waypoint, and a smooth path that passes through the waypoints without stopping. For the selected planner the model returns the total time, the peak speed, acceleration and jerk, the acceleration and jerk effort, whether the speed and acceleration limits are met, the time-scale factor that would just meet them, the path's bounding box, and 201 frames of the tracked body with its error and control force. Five scored challenges use the same model.
It is a lab for learning trajectory planning and tracking: why stopping at every waypoint costs time, how the sharing of time between segments shapes a smooth path, why feedforward lets feedback correct only small errors, and how a sampled controller can become unstable. It is not a motion-control design tool. There are no obstacles, no motor or drive models, and the body is a point mass on a frictionless floor.
The model is the calculation authority. The page sends the values you set to the model and draws only what it returns: the planned path, the reference and the tracked body at each frame, the error line, the force arrow, and the speed and acceleration charts against the limit lines. The page computes no position, speed, force or time itself.
Inputs and units
Positions are in metres (m), times in seconds (s), speeds in m/s, accelerations in m/s², the mass in kg and forces in newtons (N).
| Input | Meaning and unit |
|---|---|
| Waypoints | Up to 6 rows, each an x and y from 0 to 4 m, the floor's bottom-left corner at the origin. At least two rows must be in use. Rows are visited in row order, and blank rows are skipped. Consecutive waypoints must be at least 0.05 m apart; the first and last may coincide, which makes a closed loop. |
| Planner | Trapezoid (stop at each), Minimum jerk (stop at each) or Smooth through the waypoints (default). |
| Timing rule | Uniform, Chord length (default) or Manual shares: how the smooth planner shares its total time between the segments. |
| Total time (smooth planner) | From 1 to 60 s (default 9): the smooth planner's total time T. |
| Time shares | One share per segment, each from 0.1 to 10, or blank (a blank share counts as 1). Used with Manual shares only. |
| Speed limit | From 0.1 to 5 m/s (default 1). |
| Acceleration limit | From 0.1 to 10 m/s² (default 1.5). |
| Tracked mass | From 0.1 to 100 kg (default 2): the mass m of the body that follows the plan. |
| Force limit | From 0.1 to 1,000 N (default 20): the largest control force F. |
| Loop bandwidth | From 0.5 to 50 rad/s (default 4): ωc, the natural frequency of the feedback loop. |
| Loop damping ratio | From 0.2 to 2 (default 0.7): ζ, the damping ratio of the feedback loop. |
| Feedforward | With feedforward (default) or Feedback only. |
| Challenge number | 0 for free exploration, or 1 to 5 to load a challenge. |
| Your prediction (challenge 1) | Not yet predicted, Trapezoid first, Minimum jerk first or Smooth through first: your answer in challenge 1. |
The timing rule, total time and shares apply to the smooth planner only; the two stop-at-each planners always run at their shortest times within the limits.
Where each input is set:
- On the canvas only: the waypoints, the planner (its tabs) and your arrival prediction in challenge 1. Tap an empty spot of the floor to add a waypoint (a new one is visited last), and remove one with its x or Delete.
- Draggable on the canvas: every waypoint, in 0.01 m steps.
- In the input panel: the timing rule, total time, time shares, speed and acceleration limits, the tracked mass, force limit, loop bandwidth, damping ratio, feedforward and the challenge number.
Governing relationships
Below, v is the speed limit, a the acceleration limit, and L the length of a straight segment between two consecutive waypoints.
Trapezoid (stop at each). Along each segment the body accelerates at a, cruises at v and decelerates to rest:
peak speed vp = min(v, √(L·a)), acceleration time ta = vp / a, segment time T = L / vp + ta.
When the segment is long enough (L ≥ v²/a) this is L/v + v/a; otherwise the profile is triangular and the time is 2√(L/a). It is the fastest rest-to-rest motion within the two limits. Its acceleration effort, the integral of the squared acceleration over time, is 2·a·vp, and its jerk is unbounded because the acceleration switches instantly.
Minimum jerk (stop at each). On normalised time τ from 0 to 1 each segment follows the rest-to-rest quintic
S(τ) = 10τ³ - 15τ⁴ + 6τ⁵,
which starts and ends with zero speed and acceleration and minimises the integral of the squared jerk. Its derivatives have peaks of 15/8 (speed), 10/√3 (acceleration) and 60 (jerk). The segment time is the shortest that meets both limits, Tq = max(15L / (8v), √(10L / (√3·a))). Then the peak speed is 15L/(8Tq), the peak acceleration (10/√3)·L/Tq², the peak jerk 60L/Tq³, the acceleration effort (120/7)·L²/Tq³ and the jerk effort 720·L²/Tq⁵.
Smooth through the waypoints. The segment times h_k are T/n (Uniform), T·L_k divided by the total path length (Chord length), or T·w_k divided by the sum of the shares (Manual shares). On each segment each axis is a cubic in time (Hermite form) from waypoint to waypoint, starting and ending at rest. The velocities at the interior waypoints are chosen so the acceleration is continuous; this is a clamped cubic spline, found by solving a small tridiagonal system by forward elimination and back substitution. For fixed segment times the clamped cubic gives the least acceleration effort among smooth paths through the waypoints. On each segment the acceleration changes linearly, so its peak is at an end and is found exactly, and the jerk is constant.
The speed of the smooth path is sampled at 65 evenly spaced points of each segment. The true peak can exceed the sampled peak by at most the speed check bound, half the segment's peak acceleration times h_k / 64. The bounding box of the smooth path is found exactly from the roots of each axis's velocity; for the stop-at-each planners the path is the straight polyline through the waypoints.
Limits and the time-scale factor. A plan meets the limits when its peak speed is at most v and its peak acceleration at most a (each with a relative band of 1×10⁻⁹). Stretching every segment time by a factor c divides every speed by c and every acceleration by c², so
scale factor c = max(peak speed / v, √(peak acceleration / a)),
and c x (total time) is the time at the limits: the total time at which the same plan just meets both limits. The fastest planner is the one with the least time at the limits; a tie within 1×10⁻⁹ (relative) goes to the earlier planner in the list.
Tracking. The plan is sampled at 201 frames, from 0 to the total time Tt, with step h = Tt / 200. The body starts at rest at the first waypoint. With the errors e (reference position less body position) and ė (reference velocity less body velocity) on each axis, the proportional-derivative (PD) controller with optional feedforward commands
u = m·a_ref (feedforward only) + Kp·e + Kd·ė, with Kp = m·ωc² and Kd = 2·ζ·m·ωc.
When the commanded force's size exceeds F it is scaled down to F, keeping its direction. The force is held for one step, and the body, a double integrator, moves exactly under a constant force:
x_next = x + h·vx + (h²/2)·(ux / m), vx_next = vx + h·(ux / m), and the same for y.
Discrete stability (Jury test). Without the force limit, each axis's error evolves by a fixed 2 x 2 matrix whose characteristic polynomial is z² - tr·z + dt, with x = ωc·h, tr = 2 - x²/2 - 2ζx and dt = 1 - 2ζx + x²/2. Jury's conditions for a second-order polynomial reduce to
ζ·ωc·h < 1 and ωc·h < 4ζ,
so the sampled loop is stable exactly when both hold. A marginal loop counts as unstable. Above ωc·h = 0.5 the sampled controller no longer behaves like a continuous one.
Calculation sequence
- Check the inputs in the order of the status list under Errors and warnings. In a challenge, take the challenge's fixed values, except the inputs it leaves to you.
- Read the waypoints in use in row order and find each segment's length and the total path length.
- Plan the segments by all three planners: the trapezoid and minimum-jerk times and peaks, and the smooth planner's segment times, interior velocities, cubic segments, sampled speeds and bounding box.
- Summarise each planner (total time, peaks, efforts, scale factor and time at the limits), and find the fastest.
- For the selected planner, set the 201 frame times and the reference position, velocity and acceleration at each.
- Run the controller and the exact held-force plant for 200 steps, recording the error and the force.
- Find the tracking scalars and the stability verdict.
- Set the status and, in a challenge, score it.
Outputs and interpretation
- Total time (s) of the selected planner, the path length through the waypoints (m) and the segments in use.
- Peak speed (m/s), peak acceleration (m/s²) and peak jerk (m/s³); the jerk reads as unbounded for the trapezoid.
- Acceleration effort (m²/s³) and jerk effort (m²/s⁵): the integrals over time of the squared acceleration and jerk. Smaller means a gentler motion.
- Speed check bound (m/s): how far the smooth planner's true peak speed can exceed the sampled one (0 for the stop-at-each planners, whose speed is exact).
- Limits met (Yes or No) and the limit note: Within both limits, Over the speed limit, Over the acceleration limit or Over both limits.
- Scale factor, time at the limits (s) and the fastest planner.
- Bounding box of the path (m), inside the workspace (Yes or No) and how far it goes outside (m).
- Sample step h (s), bandwidth times step ωc·h, and sampled loop stable (Yes or No).
- Largest, RMS (over the 201 frames) and final tracking error (m).
- Share of steps force-limited (0 to 1), peak control force (N) and control energy (N² s), the integral of the squared force over time.
- Settings used: every planning and tracking input in effect.
- Frames table: for each of the 201 frames, the time, the reference and body positions, the reference speed and acceleration, the error, the force components and whether the force was limited.
- Segment table: each segment's waypoint rows, start time, duration, length, peak speed, peak acceleration and acceleration effort for the selected planner.
- Planner comparison: each planner's total time, time at the limits, peaks and efforts.
Reading the canvas: the planned path is drawn through the numbered waypoints; at Experiment the tracked body, the line from it to the reference (the error) and its control force arrow are shown at the current frame. The charts plot the reference speed and acceleration against the limit lines. A frame on a segment or phase boundary (within 1×10⁻⁹ s) belongs to the later one.
A plan over a limit, a force limit that acts, a smooth path that leaves the floor and a large tracking error are results, not warnings: they are what several examples and challenges are about.
Challenges
Each challenge fixes its own waypoints, planner settings, limits and tracking settings, except the inputs it leaves to you. The model scores each one MET (100) or NOT MET (0) and gives a one-line hint while it is not met.
- Which arrives first? Predict which planner reaches the last waypoint first when each one runs at the speed and acceleration limits. It is met when your prediction names the planner with the least time at the limits.
- Beat the clock. Reach the last waypoint within 8 s without breaking the 2 m/s speed limit or the 1 m/s² acceleration limit. It is met when the total time is at most 8 s, the peak acceleration is within the limit, and the peak speed plus the speed check bound is within the speed limit (a guarantee, not a sample).
- Share the time. With 10 s in total, share the time between the segments so that the acceleration effort is at most 2.25 m²/s³.
- Track within 2 cm. Keep the tracking error at or below 2 cm everywhere, with a force limit of 3.2 N. It is met when the largest tracking error is at most 0.02 m and the sampled loop is stable.
- Stay on the floor. Pass through all four waypoints without the smooth path leaving the 4 m workspace.
Challenge 1 holds back the answer until you predict: while your prediction reads Not yet predicted, the scale factor, the time at the limits, the fastest planner and the planner comparison are blank, and the challenge reads NOT MET with a prompt to predict. Challenge 1 fixes the planner, so the comparison cannot be read by switching planners inside it.
What to take from the lab: stopping at every waypoint is time-optimal per segment but slow overall; sharing time to suit each segment's length keeps a smooth path gentle and close to the waypoints; and feedforward supplies the force the plan needs, so feedback has only small errors to correct.
Validation and status logic
The status reads the first match in the list under Errors and warnings. Items 1 to 12 are checked on the values you entered, including inputs the selected planner does not use; the three CHECK items are checked on the values in use, and every result is shown under them. A challenge number that is not a whole number from 0 to 5 counts as no challenge.
Assumptions and limitations
- A point body on a flat, frictionless 4 m square floor; no obstacles, no motor or drive dynamics.
- Up to 6 waypoints visited in row order; the trapezoid and minimum-jerk planners move in straight lines and stop at every waypoint.
- The smooth planner's timing is yours to choose; the model reports the uniform time stretch that just meets the limits, not an optimum over the shares.
- The smooth planner's peak speed is sampled at 65 points a segment; the true peak exceeds it by at most the speed check bound.
- The controller is sampled at 200 steps over the whole plan, so a longer plan has a coarser step. The force limit scales the commanded vector; the stability verdict ignores the force limit.
- The RMS error is over the 201 frames and the control energy over the 200 held steps; the force at the last frame is shown, not applied.
- Each test has a stated band of 1×10⁻⁹ (time, length or relative); a value on a band edge can flip a reading.
Restrictions and non-computing states
Under NOT VALID every result and every table is blank, apart from the model version, the challenge status (No challenge selected, or NOT MET when a challenge is selected), a score of 0, and the goal and hint of a selected challenge.
The peak jerk and jerk effort are blank for the trapezoid, whose jerk is unbounded. Segment rows beyond the segments in use are blank.
In challenge 1, until you predict, the comparison results are held back as described under Challenges.
Errors and warnings
The status messages, first match wins (k and k2 are waypoint row numbers, or a segment number):
- NOT VALID: waypoint k needs both x and y from 0 to 4 m, or leave the row blank
- NOT VALID: place at least two waypoints on the course
- NOT VALID: waypoints k and k2 are closer than 0.05 m; move them apart or clear one of the rows: the first pair of consecutive waypoints in row order that is too close.
- NOT VALID: choose the planner from the list
- NOT VALID: choose the timing rule from the list
- NOT VALID: the total time must be from 1 to 60 s
- NOT VALID: time share k must be a number from 0.1 to 10, or leave it blank (a blank share counts as 1)
- NOT VALID: the speed limit must be from 0.1 to 5 m/s and the acceleration limit from 0.1 to 10 m/s2
- NOT VALID: the mass must be from 0.1 to 100 kg and the force limit from 0.1 to 1000 N
- NOT VALID: the loop bandwidth must be from 0.5 to 50 rad/s and the damping ratio from 0.2 to 2
- NOT VALID: choose the feedforward setting from the list
- NOT VALID: the challenge number must be a whole number from 0 to 5, and the arrival prediction one of the listed choices
- CHECK: the sampled tracking loop is unstable at this sample step, so the error grows until the force limit holds it; lower the bandwidth or shorten the plan: the Jury conditions above fail.
- CHECK: the loop bandwidth is high for the sample step (bandwidth times step above 0.5), so the sampled controller no longer behaves like a continuous one; lower the bandwidth or shorten the plan
- CHECK: segment k lasts less than 0.1 s, so its speeds and accelerations are extreme and the frames cannot show it; give that segment more time: smooth planner only.
- OK
References
This laboratory is an educational model of trajectory planning and tracking for a point body on a flat floor. It compares three closed-form plans: a trapezoidal velocity profile, the fastest rest-to-rest motion within a speed and an acceleration limit; the rest-to-rest quintic of the minimum-jerk model; and a clamped cubic spline in time through the waypoints, whose interior velocities come from a small tridiagonal system. A body of given mass then follows the plan under a sampled proportional-derivative controller with optional feedforward and a force limit, moving exactly as a double integrator under a held force, and Jury's test gives the closed-form stability condition of the sampled loop.
Sources and further reading
- The trapezoidal (ramp) velocity profile, its acceleration, constant-speed and deceleration phases, and the instant switches of acceleration that make its jerk unbounded: B. Schweber, Basics of motion-control profiles, Part 2: Ramp profiles, EE World, 2021. EE World
- The minimum-jerk model of point-to-point movement: T. Flash and N. Hogan, "The coordination of arm movements: an experimentally confirmed mathematical model", Journal of Neuroscience 5(7), 1985, pages 1688-1703. PubMed Central
- Cubic spline interpolation with clamped end slopes, solved as a tridiagonal linear system: Stanford University, Math 114 course notes, Spline interpolation. Stanford University
- The stability of sampled-data systems and Jury's test for a second-order characteristic polynomial: K. Iqbal, Introduction to Control Systems, Engineering LibreTexts, section 7.4, Stability of Sampled-Data Systems. LibreTexts
- Trajectory generation, time scaling and motion control with feedforward and feedback: K. M. Lynch and F. C. Park, Modern Robotics: Mechanics, Planning, and Control, Cambridge University Press, 2017, chapter 9, Trajectory Generation, and chapter 11, Robot Control. Northwestern University
No table or text from any of these publications is reproduced, and none of their authors or publishers has reviewed or endorses this laboratory.
This laboratory is for learning. Its results hold only within its stated ranges; it models a point mass with no motor, drive or friction, so it is not a basis for tuning a real machine.
Links were current at build time. If one moves, search the document's title rather than reporting a broken link as a defect.
Found a problem, or have an idea?
Tell us if a result looks wrong, a label is unclear, or something is missing. We read every message.