Professional Documents
Culture Documents
Mechatronic Systems
Input Output
Displacement Force
Spring Mass System
Source:Youtube (Lesics)
Example – Thermal Power Plant
(contd.)
https://www.kepco.co.jp/english/energy/fuel/thermal_power/shikumi/index.html
Steady State and Dynamic Operation of
Systems
Steady-state operation: system variables are assumed to remain constant in time
Dynamic operation: Behaviour as a function of time is important
Real Canal
Scaled physical model of flood control design
Source: Youtube (Stantech)
Mathematical Models
Developed based on mathematical equations and computer simulations
Physical laws and boundary conditions are used
For instance, we can model how the aircraft responds to pilot input
command signals during test maneuvers
Such a model would not have prediction capability that are not modeled
explicitly
a scaled physical model may give an idea of aerodynamic heating even if it
is not the primary intent of the model
An Example of
Mathematical Model
Anguilliform-inspired Robot
Developed @IITP
Mathematical
Model
𝑥ሷ 𝐶𝑀
𝑎𝑇 μ𝑛 𝑆𝜃2 −𝑎𝑇 μ𝑛 𝑆𝜃 𝐶𝜃 −𝑎𝑇 μ𝑛 𝑆𝜃 𝐶𝜃 −𝑎𝑇 μ𝑛 𝑆𝜃2 𝑉𝑥𝑎 ሶ 𝑎𝑇 𝑓𝐷𝑥
= −𝑀𝑝 𝑀𝑘 − 𝑀𝑝 𝑎 𝜃 + 𝑀𝑝
−𝑎𝑇 μ𝑛 𝑆𝜃 𝐶𝜃 𝑇
𝑎 μ𝑛 𝑆𝜃 2
𝑎𝑇 μ𝑛 𝐶𝜃2 𝑇 𝑉
𝑎 μ𝑛 𝑆𝜃 𝐶𝜃 𝑦 𝑎𝑇 𝑓𝐷𝑦
where 𝑀𝑘 =
Mathematical Model
𝑙𝐾 𝑇 (𝐶𝜃 𝜃ሶ 2 + 𝑆𝜃 𝜃ሷ )
𝑙𝐾 𝑇 (𝑆𝜃 𝜃ሶ 2 − 𝐶𝜃 𝜃ሷ )
, 𝑀 𝑝 =
𝑚11 𝑚12
𝑚21 𝑚22 =
2ሶ 𝑇 𝑇 −1 𝑇 𝑣𝑐 = 𝑣𝑐,𝑥 , 𝑣𝑐,𝑦 =
𝑁𝑚 𝑎𝑇xμCM
x =+[𝜃, 𝑆, 𝜃,
𝑎
𝑛 𝜃 CM xሶ ]
−𝑎 μ 𝑆
𝑛 𝜃 𝜃𝜏𝐶 =𝑎[𝜏 1 , … , 𝜏𝑁−1 ]
[𝑠𝑐 cos 𝜃𝑐 , 𝑠𝑐 cos 𝜃𝑐 ]
𝑎𝑇 μ𝑛 𝑆𝜃 𝐶𝜃 𝑎 𝑁𝑚 + 𝑎𝑇 μ𝑛 𝐶𝜃2 𝑎
x: Robot State 𝜏: Torque Vector 𝑣𝑐 : Flow Velocity
𝑎 = 1 𝜃…:1Joint𝑇
ℜ𝑛 , 𝑓𝐷𝑥 (𝐶𝜏𝑓𝑁−1
∈ Angle , 𝐶𝐷: )Torque
and 𝑓𝐷𝑦atare − 1)𝑡ℎ joint
(𝑁 effects of drag forces𝑠𝑐 : Flow
in x Magnitude
and y directions and
xCM : Robot Coordinates 𝑁:𝑥Number ሶ = 𝑓 of 𝜏,
(𝑥, links
𝑣 ; 𝑔 , 𝑒) 𝜃𝑐 : Flow
𝑒 = [CDirection
, 𝐶 , 𝐶 , 𝐶 ]𝑇
is a sum of linear and non-linear drag forces. 𝑐 𝑝 𝑓 𝐷 𝐴 𝑚
𝜃ሷ = −𝑀𝜃 [𝐷𝑇 𝜏 − 𝑊𝜃 𝜃ሶ 2 − 𝑉𝜃 𝜃ሶ − Λ3 𝜃ሶ 𝜃ሶ − 𝐾𝐷𝑥 𝑓𝐷𝑥 − 𝐾𝐷𝑦 𝑓𝐷𝑦 ]
𝑐𝑖𝑟𝑐𝑙𝑒 x𝐶𝑀,𝑔 2𝑇 𝑒:
𝑇 𝑆Environment
2 𝜇]𝑇𝐾 𝑇Parameters
𝑆𝜃𝜃=diag([sin
𝑀 2
= 𝐽 + 𝑚𝑙 𝜃𝑆1𝜃 𝑉𝑆 …𝜃sin+ 𝑚𝑙𝜃𝑁 ] 𝐶)𝜃and
𝑉𝐶𝜃 𝐶 Λ1 + 𝑙 2 𝜇𝑛𝑔𝐾𝜃11=
+𝜃 =diag([cos 𝐾… [𝐴,cos
𝜃 +
𝜔, 𝑙𝜃
𝛽, 𝑁𝑘 𝑛 )
]𝑇,
2 𝐾𝐾 𝐶=𝜃 𝑨𝑇 𝑫𝑫𝑇 −1 𝑫
𝑜𝑓
𝑊𝜃 = 𝑚𝑙 2 𝑆𝜃 𝑉𝑆𝜃 − 𝑚𝑙 ϵ 2 𝐶 𝑉𝑆 + 𝑙 2 𝜇 𝐾 𝐾C𝑓𝑇:𝐶Drag +
𝑝 2Coefficient
𝑙 𝜇 𝐾 𝐾 𝑇 𝑆 𝜃along longitudinal direction
𝜃 𝜃 𝑛 1 𝜃 𝑛 2 𝜃
𝑎𝑐𝑐𝑒𝑝𝑡𝑎𝑛𝑐𝑒 𝑎 𝑎 𝐶 : Drag Coefficient along transverse direction
𝑉𝜃 = Λ11 1− 𝑙 𝜇𝑛 𝐾2 𝑉𝑥 − 𝑙𝜇𝑛 𝐾21𝑉𝑦11 𝐷
𝑔𝑝 : Gait Parameters
A == 𝐴 ⋱+⋱𝜇 𝐴 𝑎𝑎𝑡 𝑚D 𝑆 =𝐶 − ⋱𝑚 ⋱ 2 𝐶𝑡𝐴 : Added mass coefficient
𝐾 1 1 𝑛 1 12 𝜃 𝜃 11 𝜃 𝑆 − 𝜇 𝑛 2𝐴 𝑎𝑎
𝐴: (𝑚22 𝑆𝜃 𝐶𝜃 − 𝑚21 𝑆𝜃2 )
Amplitude
𝐵1 1 𝑁−1𝑡 𝑁 1 −1 2𝑁−1 𝑁 𝐶𝑡𝑀 : Added inertia coefficient
𝐾2 = 𝐴2 − 𝜇𝑛 𝐴1 𝑎𝑎 𝜃 𝑟𝑒𝑓 𝑚11 𝑆𝜃 𝐶𝜃 − 𝑚12 𝑆𝜃 − 𝜇𝑛 𝐴2𝜔: 𝑚21 𝑆𝜃 𝐶𝜃 − 𝑚22 𝐶𝜃2
𝑎𝑎 Frequency
𝑅
𝐴1 = 𝑆𝜃 𝐾 𝑆x𝜃2𝐶𝑀+ 𝐶𝜃 𝐾 𝑆𝜃 𝐶𝜃 , 𝐴2 = 𝑆𝜃 𝐾 𝑆𝜃 𝐶𝜃𝑎 + 𝐶𝜃 𝐾𝐶 𝛽:𝜃2 Phase
𝑎
Shift
μ𝑛 =F(𝐶𝐴ҧ ) is the addedሷ mass ሶ factor, ሶ 𝑉𝑥ሶ (𝑣𝑘𝑐,𝑥:) High-Level
ΩFluid and 𝑉𝑦 (𝑣𝑐,𝑦Controller
)
𝜃
Torques 𝑇 = Λ1 𝜃 − Λ2 𝜃 − Λ3 𝜃|𝜃| 𝜃
Λ 𝐶 , Λ 𝐶 and Λ 𝐶 , 𝐽 is the moment of inertia Gain
of the links
1 𝑀 1 𝑓 1 𝑓
Mathematical Model
n−i
Dynamics of the robot: Controller:
𝛼= A sin(𝜔t + (i−1)𝛽) + 𝛾
n−1
𝜏 = 𝑘𝑝 e + k d 𝑒ሶ where e = 𝛼 − 𝛼 ∗
𝑥ሶ = 𝑓 (𝑥, 𝜏, 𝑣𝑐 ; 𝑔𝑝 , 𝑒)
Closed form solution of the equation cannot be found, hence we discretize the
dynamics and the controller
𝐵
𝜃𝑟𝑒𝑓
𝑅
x𝐶𝑀
Ω 𝜃ҧ
13
Controller Circle of Goal
Start
acceptance
𝑥𝐶𝑀 𝑥𝐶𝑀,𝑔
14
Controller
Start Goal
𝑥𝐶𝑀 𝑥𝐶𝑀,𝑔
15
State-Determined Systems
Expressed as ordinary differential equations in terms of state variables
State of a system is a set of variables that along with the input and dynamics equation can
predict the future state and output of the system
Example of spring mass damper system
𝑑 2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝑀 +𝑏 + 𝑘𝑥(𝑡) = 𝑢(𝑡)
𝑑𝑡 2 𝑑𝑡
𝑑𝑥 𝑡
Let, 𝑥1 𝑡 = 𝑥 𝑡 and 𝑥2 𝑡 =
𝑑𝑡
0ሶ 1 𝑥1 0
𝑥ሶ 1 𝑘 𝑏 1 𝑢
= 𝑥2 +
𝑥ሶ 2 − −
𝑀 𝑀 𝑀
If we are
In general, for linear systems interested in
𝑋ሶ = 𝐴𝑋 + 𝐵𝑈 displacement
And for nonlinear systems 𝑥(𝑡) as output
𝑋ሶ = 𝑓(𝑋, 𝑢) then
Input Output 𝑌(𝑡) = 𝑥(𝑡)
Variables Dynamic System, 𝑆
Variables
𝑈 State variable, 𝑋
𝑌
State-space formulation is often used for mathematical modeling
Applications of Modeling
Analysis
Given present state 𝑋, action 𝑈, model 𝑆, determine future output 𝑌
Identification
Given history of actions 𝑈 and outputs 𝑌 obtained by experiments,
determine model 𝑆 and state variable 𝑋
Synthesis
Given action 𝑈 and desired output 𝑌, find system 𝑆 such that 𝑈
acting on 𝑆 will generate output 𝑌
Control
Given system 𝑆, design control system that uses 𝑌 obtained from
sensors to determine control input 𝑈 to make the system respond in
desired way meeting prespecified performance criteria
Multiport Systems and Bond Graphs
Engineering Multiports
When two components are joined together physically
Two complementary variables are simultaneously constrained to be equal for them
Ports are places at which subsystems can be interconnected
Power flows via ports
Systems with one or more ports are called multiports
𝑄 𝑃
Hydraulic
𝑖 Electric Motor 𝜏 2 𝜔2 Pump
𝜔 𝜏 𝜔1 𝜏1 𝜏 𝜔
𝑄
𝑒
2-port systems
3-port system
Power Variables
Variables that are forced to be identical when two multiports are connected
are called power variables
Product of power variables is the instantaneous power flowing between the two
multiports
𝑄 𝑃
Example
Hydraulic
𝑖 Electric Motor Pump
𝜔 𝜏 𝜏 𝜔
𝑄
𝑒
𝑡 𝑑𝑞 𝑡
𝑝 𝑡 = න 𝑒 𝑡 𝑑𝑡 = 𝑝0 + න 𝑒 𝑡 𝑑𝑡 = 𝑓 𝑡 , 𝑑𝑞 = 𝑓𝑑𝑡
𝑡0
𝑑𝑡
Modeling of Systems using Power and Energy Variables
Any system can be modeled using only four variables
power (effort 𝑒 and flow 𝑓) and
energy (momentum 𝑝 and displacement 𝑞)
Generalized Mechanical Mechanical Hydraulic Port Electrical Port
Variables Translation Port Rotation Port
Effort, 𝑒 Force, 𝐹 (N) Torque, 𝜏 (N-m) Pressure, 𝑃 (N/m2) Voltage, e, (V or N-
m/C)
Flow, 𝑓 Velocity, 𝑉 (m/s) Angular velocity, Volume flow rate, 𝑄 Current, I, (A or
𝜔 (rad/s) (m3/s) C/s)
Momentum, Momentum, 𝑃 (N-s) Angular Pressure momentum, Flux linkage, 𝜆(V-s)
𝑝 Momentum, 𝑝𝜏 𝑝𝑝 (N-s/m2)
(N-m-s)
Displacement, Displacement, 𝑋 (m) Angle, 𝜃 (rad) Volume, 𝑉 (m3) Charge, q (A-s)
𝑞
Power, 𝑃 𝐹(𝑡)𝑉(𝑡) (N-m/s) 𝜏 t 𝜔(𝑡) (Nm/s) 𝑃(𝑡)𝑄(𝑡) (Nm/s) e 𝑡 𝑖(𝑡) (V-A or W
or Nm/s)
Energy, 𝐸 𝑥𝑑𝐹 , ( 𝑃𝑑𝑉 N- 𝜃𝑑𝜏 , 𝑉𝑑𝑃 , ) 𝑝𝑝𝑑𝑄 𝑞𝑑𝑒 , ( )𝜆𝑑𝑖 V-A-
m) ( ) 𝑢𝑎𝑡𝑝𝑑𝜔 N- (N-m) s, or W-s or N-m)
m)
Bond Graph Representation
effort
flow
effort
flow
Example – Pump, Motor, Battery
Rotational inertia
Effort Source and Flow Source
Voltage supply, pressure source, vibration shakers, constant-
flow systems
Basic 2-port Elements Ideal rigid lever
Gear pair
Constitutive laws; 𝑒1 = 𝑚𝑒2 and 𝑚𝑓1 = 𝑓2
where 𝑚 is called transformer modulus
Hydraulic ram
Gyrator
A gyrator’s constitutive laws are 𝑒1 = 𝑟𝑓2 and 𝑟𝑓1 = 𝑒2 where r
is the gyrator modulus
Electrical Gyrator
Mechanical Gyrator
Voice Coil
Relation between Gyrator and
Transformer
Two gyrators cascaded together gives a transformer
Modulated Transformer and Modulated Gyrator
Modulated Modulated
transformer gyrator
Example
Electrical Hydraulic
Mechanical
Relationship of 0 and 1 junctions to Physical Laws
System Bond Graph Symbol Physical Laws
Transformer
Gyrator
Causal Form for 3-Ports
Thermal Systems as Pseudo-bond Graph
Thermal systems often represented analogous to electric circuits
Temperature as voltage
Heat flow as current
With above analogy thermal resistor, capacitor, sources can be imagined –
however there is no thermal inertia
Temperature is effort and heat flow is flow
One problem though is that temperature*heat flow ≠ power
Bond graph in which effort*flow is not power is called pseudo-bond
graph
Cannot be coupled to normal bond graph using power variables
Where,
l: thickness, k: thermal
conductivity, A:area
u is internal energy per unit mass